Chris Fraser | 中国早期思想中墨家与荀子的知识论

注：本文为 “中国早期思想中的知识论” 相关合辑。
英文引文，机翻未校。
如有内容异常，请看原文。

---

Knowledge and Error in Early Chinese Thought

中国早期思想中的知识与错误

Chris Fraser

Published online: 31 March 2011
Springer Science+Business Media B.V. 2011
Chris Fraser (Department of Philosophy, University of Hong Kong, 312-315 Main Building, Pokfulam Road, Hong Kong, Hong Kong
e-mail: [e-mail]
Chris Fraser

Abstract

摘要

Drawing primarily on the Mòzǐ and Xúnzǐ, the article proposes an account of how knowledge and error are understood in classical Chinese epistemology and applies it to explain the absence of a skeptical argument from illusion in early Chinese thought. Arguments from illusion are associated with a representational conception of mind and knowledge, which allows the possibility of a comprehensive or persistent gap between appearance and reality. By contrast, early Chinese thinkers understand mind and knowledge primarily in terms of competence or ability, not representation. Cognitive error amounts to a form of incompetence. Error is not explained as a failure to accurately represent the mind-independent reality due to misleading or illusory appearances. Instead, it can be explained metaphorically by appeal to part-whole relations: cognitive error typically occurs when agents incompetently respond to only part of their situation, rather than the whole.

本文主要以《墨子》和《荀子》为依据，阐述了中国古典认识论中对知识与错误的理解，并以此解释中国早期思想中为何没有出现来自幻觉的怀疑论论证。来自幻觉的论证与心灵和知识的表征性概念相关，这种概念承认表象与实在之间可能存在全面或持久的鸿沟。相比之下，中国早期思想家主要从能力或胜任力的角度理解心灵和知识，而非表征。认知错误是一种无能的表现。错误并非被解释为由于误导性或虚幻的表象而未能准确表征独立于心灵的实在，相反，它可以通过部分与整体的关系进行隐喻性解释：认知错误通常发生在主体未能恰当地回应其处境的整体，而只回应了部分之时。

Keywords Epistemology. Mind . Knowledge . Illusion . China
关键词 认识论；心灵；知识；幻觉；中国

1 Introduction

1 引言

We are all familiar with perceptual illusion. Walking home at twilight, we see a person resting on a park bench ahead, only to discover a few steps later that the “person” is actually a small tree by the bench. Washing our hair in the shower, we hear the phone ring, only to check the phone log afterward and find no missed calls. Such everyday experiences inspire the traditional skeptical argument from illusion. Illusions confirm that our senses sometimes deceive us, a simple form of the argument runs. Hence sense perception is too unreliable to yield knowledge, for perhaps it always deceives us, or at least we cannot be sure when it does and when not.

我们都熟悉知觉幻觉。暮色中步行回家时，我们看到前方公园长椅上有个人在休息，走几步后却发现那个 “人” 其实是长椅旁的一棵小树。在淋浴时洗头，我们听到电话响了，事后查看通话记录却发现没有未接来电。这些日常经历催生了传统的来自幻觉的怀疑论论证。这种论证的一种简单形式认为，幻觉证实了我们的感官有时会欺骗我们。因此，感官知觉太不可靠，无法产生知识，因为它可能总是在欺骗我们，或者至少我们无法确定它何时在欺骗、何时没有。

The argument from illusion, in its various guises, presents a perennial challenge for epistemologists.1 Versions of the argument have had a central role in motivating skepticism in both its ancient and modern forms. The argument drives the foundationalist search for an absolutely certain starting-point for beliefs, and it contributes to the transcendental anxiety that epistemological coherentism or holism might allow our system of beliefs to lose touch with independent reality, such that empirical thought degenerates into mere “frictionless spinning in a void” (McDowell 1996: 11).
各种形式的来自幻觉的论证，对认识论者构成了长期的挑战。1 这种论证的不同版本在推动古代和现代形式的怀疑论方面都发挥了核心作用。它促使基础主义者去寻找信念的绝对确定的起点，也加剧了一种先验焦虑：认识论的融贯论或整体论可能会使我们的信念系统与独立的实在失去联系，从而使经验思想堕落为纯粹的 “无摩擦的空转”（麦克道威尔 1996：11）。

1 For convenience, I will refer to “the argument from illusion” in the singular, though in fact there are a variety of arguments from illusion.
1 为方便起见，我将用单数形式提及 “来自幻觉的论证”，尽管实际上存在多种来自幻觉的论证。

An intriguing aspect of early Chinese epistemology is that the argument from illusion plays no comparable role, even though perceptual illusion is explicitly considered in two of the most important pre-Hàn epistemological texts, the Mohist “Dialectics” (Mòbiàn 墨辯) and Book 21 of the Xúnzǐ 荀子，“Resolving Obscuration (Jiěbi 解敝).”2 Neither the Mohists nor Xúnzǐ regards illusion as grounds for skepticism or a threat to knowledge. To the contrary, an implicit theme of Xúnzǐ’s discussion is that familiar cases of illusion show that perceptual knowledge is reliable, not unreliable, since agents who employ their cognitive abilities properly will not be misled in such situations.
中国早期认识论的一个有趣方面是，来自幻觉的论证并未发挥类似作用，尽管汉代以前最重要的两部认识论文献 —— 墨家《墨辩》和《荀子》第 21 篇《解蔽》—— 都明确讨论了知觉幻觉。2 墨家和荀子都不认为幻觉是怀疑论的根据或对知识的威胁。相反，荀子的论述中隐含着一个主题：常见的幻觉案例表明知觉知识是可靠的，而非不可靠，因为恰当运用认知能力的主体在这类情况下不会被误导。

2 The Mohist “Dialectics” is a modern label for the numerous short texts collected together in books 40– 45 of the Mòzǐ, which focus on language, knowledge, argumentation, ontology, and ethics, among other topics. Throughout the article, I will refer to these texts collectively as the “Dialectics.” They probably date from between 320 and 250 B.C.E. “Resolving Obscuration” is likely from the first half of the third century B.C.E.
2 墨家 “辩经” 是一个现代标签，指的是《墨子》第 40 至 45 篇中收集的众多短文，这些短文主要讨论语言、知识、论证、本体论和伦理等主题。在整篇文章中，我将这些文本统称为 “辩经”。它们的年代可能在公元前 320 年至公元前 250 年之间。《解蔽》则可能出自公元前 3 世纪上半叶。

The aim of this essay is to sketch how knowledge and error are understood in early Chinese thought and by so doing to explain why illusion raises no deep skeptical worries for early Chinese thinkers, none of whom even considers a skeptical argument from illusion. The explanation I propose turns on two main claims. The first-which I will merely sketch, without defending in detail-is that the power of the argument from illusion rests on what I will call a “representational” conception of mind and knowledge. On such a conception, perceptual illusion occurs when the senses misrepresent the state of the world. Such misrepresentation raises the possibility of a lasting or comprehensive gap between how things seem to the agent and how they really are. This possibility generates the skeptical force of the argument from illusion.
本文旨在概述中国早期思想中对知识与错误的理解，并以此解释为何幻觉没有引起中国早期思想家的深层怀疑论担忧 —— 他们中没有一人考虑过来自幻觉的怀疑论论证。我提出的解释基于两个主要观点。第一个观点 —— 我仅作概述，不做详细辩护 —— 是，来自幻觉的论证的力量源于我所谓的心灵和知识的 “表征性” 概念。根据这种概念，当感官歪曲世界状态时，就会产生知觉幻觉。这种歪曲使得事物对主体的显现与其实际所是之间可能存在持久或全面的鸿沟，而这种可能性正是来自幻觉的论证具有怀疑论力量的原因。

The second claim is that early Chinese thinkers understand mind and knowledge mainly in terms of competence or ability, not representation.3 For them, the major function of the xīn 心，or “heart,” is to guide action by discriminating different kinds (lèi 類) of things, thus triggering skilled responses to them.4 Knowledge is understood primarily as a reliable ability to discriminate things and respond to them appropriately- minimally, by predicating appropriate general terms of them. To know is thus to be able to perform certain tasks competently. Cognitive error is a form of incompetence or ineptitude in discriminating and responding to things. Since incompetent, inept performances nevertheless do not cease to be performances-they genuinely engage with the world, albeit inappropriately-error is not associated with the possibility of the agent somehow failing to connect with mind-independent reality. Hence error due to illusion raises no skeptical worries. Conditions in which illusions occur make correct discrimination more difficult, but a competent agent knows how to cope with them. Should an agent be misled by an illusion, the explanation is not that she has mistaken appearance for reality, but that she has responded to only some features of her circumstances, overlooking other relevant features. She has attended to part of her situation at the expense of a correct response to the whole. For early Chinese epistemologists, I propose, the distinction between part and whole takes the place of that between appearance and realityin explaining perceptual error and cognitive error more generally.5
第二个观点是，中国早期思想家主要从胜任力或能力的角度理解心灵和知识，而非表征。3 对他们而言，“心” 的主要功能是通过区分不同种类（类）的事物来指导行动，从而引发对这些事物的熟练回应。4 知识主要被理解为可靠地辨别事物并对其做出恰当回应的能力 —— 至少是能将恰当的一般术语归属于它们的能力。因此，“知” 就是能够胜任地完成某些任务。认知错误是在辨别和回应事物时的一种无能或不熟练的表现。尽管无能、不熟练的行为仍然是行为 —— 它们真实地与世界打交道，只是方式不当 —— 但错误并不涉及主体在某种程度上未能与独立于心灵的实在相联系的可能性。因此，由幻觉导致的错误不会引发怀疑论担忧。产生幻觉的条件使正确的辨别更加困难，但有能力的主体知道如何应对这些条件。如果主体被幻觉误导，原因并非她将表象误认为实在，而是她只回应了自身处境的某些特征，忽略了其他相关特征。她关注了处境的一部分，却未能对整体做出正确回应。我认为，在中国早期认识论者看来，在解释知觉错误和更一般的认知错误时，部分与整体的区分取代了表象与实在的区分。5

3 Here I am developing a line of interpretation introduced by Hansen (see Hansen 1983: 64–65). To clarify, the claim here is not that early Chinese theorists do not ascribe representational functions to thought and language or understand them as having representational content. It is that their theoretical approach to language, thought, and knowledge centers on practical ability or competence, which provides the basis for their explanation of representational content. On these points, see Fraser 2009a, especially Sections 4.2 and 6.
3 在此，我发展了汉森提出的一种解释思路（见汉森 1983：64–65）。需要澄清的是，这里的观点并非中国早期理论家不认为思想和语言具有表征功能，或不将它们理解为具有表征内容。而是说，他们对语言、思想和知识的理论探讨以实践能力或胜任力为核心，这为他们解释表征内容提供了基础。关于这些观点，参见弗雷泽 2009a，特别是第 4.2 节和第 6 节。

4 The xīn is conceived of as a physical organ in the center of the body that combines cognitive, conative, and affective functions. Because early Chinese texts do not distinguish between reason and passion or the mind and heart, the word “xīn” is sometimes translated as “heart-mind.”
4 “心” 被视为身体中央的一个器官，兼具认知、意动和情感功能。由于中国早期文献没有区分理性与情感、心灵与心脏，“心” 一词有时被译为 “heart-mind”（心 - 灵）。

5 Moreover, as Jane Geaney correctly notes, in early Chinese thought, “there is no sign of the view that sense discriminations must be transcended in favor of a super-sensible form of knowing,” such as knowledge of Platonic forms or the Parmenidean unity (Geaney 2002: 13). The absence of such a view, typically associated with the metaphysical belief that the changing world of sense perception is in some sense unreal or does not reveal the ultimate or fundamental nature of reality, may be a further factor explaining early Chinese thinkers’ lack of interest in arguments from illusion.
5 此外，正如简・吉尼正确指出的，在中国早期思想中，“没有迹象表明存在这样一种观点，即感官辨别必须被超越，以支持一种超感官的认知形式”，例如对柏拉图式理念或巴门尼德式统一性的知识（吉尼 2002：13）。这种观点的缺失 —— 它通常与这样一种形而上学信念相关，即变化的感官知觉世界在某种意义上是不真实的，或者不能揭示实在的终极或根本性质 —— 可能是解释中国早期思想家对来自幻觉的论证缺乏兴趣的另一个因素。

In what follows, I first briefly explain what I mean by a representational conception of mind and knowledge and how it motivates the argument from illusion. I then review the conception of knowledge in the earliest Chinese texts that explicitly address epistemological issues-the core chapters of the Mòzǐ 墨子 - showing how knowledge is linked to ability or competence. Next, I explain how the later Mohist “Dialectics” develops this conception of knowledge, particularly with respect to perception, identifies potential causes of error, including illusion, and provides the resources for an intriguing, non-representational explanation of error, including perceptual error. I then turn to the second major early Chinese treatment of epistemological issues, Books 21 and 22 of the Xúnzǐ. I first summarize Xúnzǐ’s theory of knowledge, including perceptual knowledge, and then examine his account of cognitive error, including perceptual error and illusion. Finally, I sketch the form that skepticism takes in the early Chinese intellectual milieu. Throughout the discussion, I will highlight the respects in which the treatment of knowledge in different early Chinese texts tends to overlap and thus appears to reflect a largely shared underlying theoretical framework. I will survey the epistemological views in these texts in some detail, partly to clarify them for non-specialists and partly to illustrate the high degree of coherence between the textual details and the overarching interpretation proposed.
接下来，我首先简要解释我所说的心灵和知识的表征性概念是什么，以及它如何引发来自幻觉的论证。然后，我考察中国最早明确探讨认识论问题的文献 ——《墨子》的核心篇章中对知识的理解，展示知识如何与能力或胜任力相关联。接着，我将说明后期墨家 “辩经” 如何发展这种知识概念，特别是在知觉方面，识别包括幻觉在内的错误的潜在原因，并为包括知觉错误在内的错误提供一种有趣的、非表征性的解释。之后，我转向中国早期处理认识论问题的另一部主要文献 ——《荀子》的第 21 篇和第 22 篇。我首先概述荀子的知识理论（包括知觉知识），然后考察他对认知错误（包括知觉错误和幻觉）的解释。最后，我概述中国早期思想环境中怀疑论的表现形式。在整个讨论中，我将强调不同中国早期文献在对知识的处理上存在的重叠之处，这些重叠似乎反映了一个大体上共有的潜在理论框架。我将较为详细地考察这些文献中的认识论观点，部分是为了向非专业人士澄清这些观点，部分是为了说明文本细节与所提出的总体解释之间高度的一致性。

2 Representation and Illusion

2 表征与幻觉

The argument from illusion moves from the occurrence of perceptual error in isolated instances to the conclusion that perception might be wholly unreliable or that we are unable to tell in any particular cases whether it is reliable or not. For the argument to have any force, it must presuppose a conception of mind and knowledge on which the possibility of frequent, widespread disparity, or even a comprehensive gap, between thought and the world is intelligible. I suggest that a general feature of conceptions of mind and knowledge that render such a disparity possible and intelligible is that they appeal to a type of indirect, representational structure to explain the relation between the epistemological subject and the world.
来自幻觉的论证从孤立的知觉错误案例出发，得出结论：知觉可能完全不可靠，或者我们无法在任何特定情况下判断它是否可靠。要使这种论证具有任何力量，它必须预设一种心灵和知识的概念，根据这种概念，思想与世界之间可能存在频繁、广泛的不一致，甚至全面的鸿沟，这一点是可理解的。我认为，使这种不一致成为可能且可理解的心灵和知识概念，其一个普遍特征是诉诸一种间接的、表征性的结构来解释认识主体与世界之间的关系。

This suggestion is not a historical hypothesis. The claim is not that philosophers throughout history who have deployed versions of the argument from illusion or found them worrying each articulated or linked it to a conception of mind centered on an explicit notion of representation. Rather, the suggestion is an attempt to identify, in general, structural terms, an implicit feature of conceptions of mind or knowledge that explains how the everyday occurrence of illusion might motivate skepticism. The claim is that this structural feature is a conception of the mind as having direct access only to some intermediate entity that is in some sense “internal” to thought-such as ideas, impressions, phainomena, appearances, sense data, experiences, or a conceptual scheme-whose function is to represent or indicate how the “external,” thought-independent world is. Knowledge then lies (partly) in holding, on the basis of such entities, cognitive attitudes that accurately represent or correspond to the world. For my purposes, the crucial feature of such a conception of mind and knowledge is that, because the intermediate entities are merely gobetweens, not features of the world itself, they inherently stop short of the world. There is thus a permanent possibility that the intermediaries have misrepresented how things are on the other side.6
这一观点并非一个历史假设。它并非主张历史上所有运用过来自幻觉的论证版本或认为其令人担忧的哲学家，都明确阐述了这种论证与以明确表征概念为核心的心灵概念之间的联系。相反，这一观点试图从总体上、从结构层面识别心灵或知识概念的一个隐含特征，这个特征解释了日常幻觉的发生如何可能引发怀疑论。这种结构特征是：心灵只能直接通达某些在某种意义上是思想 “内在的” 中间实体 —— 例如观念、印象、现象、表象、感觉材料、经验或概念图式 —— 这些实体的功能是表征或指示 “外在的”、独立于思想的世界的状态。知识（部分地）在于基于这些实体持有准确表征或对应于世界的认知态度。就我的目的而言，这种心灵和知识概念的关键特征是，由于这些中间实体仅仅是中介，而非世界本身的特征，它们本质上与世界相隔绝。因此，这些中介可能永久地歪曲了另一边事物的状态。6

6 McDowell offers a contemporary account of skeptical worries arising from the argument from illusion that converges with some of my suggestions (see McDowell 1998). McDowell links skeptical worries about empirical thought collapsing into “frictionless spinning in a void” to the idea that the exercise of conceptual capacities-as exercises of Kantian spontaneity-might be unconstrained by the world (see McDowell 1996). This suggestion hooks up with my own insofar as a comprehensive gap between our scheme of representations and the world would be one cause of a lack of “constraint” or “friction.”
6 麦克道威尔对来自幻觉的论证所引发的怀疑论担忧提供了一种当代解释，这种解释与我的一些观点相契合（见麦克道威尔 1998）。麦克道威尔将对经验思想堕落为 “无摩擦的空转” 的怀疑论担忧，与这样一种观点联系起来：概念能力的运用 —— 作为康德式自发性的运用 —— 可能不受世界的约束（见麦克道威尔 1996）。我的观点与他的观点相关联之处在于，我们的表征图式与世界之间的全面鸿沟，可能是 “约束” 或 “摩擦” 缺失的一个原因。

7 Not all theories that explain mental states in terms of their representational content do so by means of such an indirect structure, of course. Approaches such as those of Dretske (see Dretske 1981) or Millikan (see Millikan 1987) assign no role to representational intermediaries and may converge in some respects with the ancient Chinese views I explore below.
7 当然，并非所有根据表征内容解释心理状态的理论都借助这种间接结构。德雷斯克（见德雷斯克 1981）或米利肯（见米利肯 1987）等人的理论没有给表征中介分配任何角色，并且在某些方面可能与我下面探讨的中国古代观点相契合。

On such a conception, cognitive error-false belief-is in effect a failure of accuracy or correspondence between the subject’s representations of the world and how the world actually is. Illusion is a species of such error that occurs when a subject’s perceptual experience is indistinguishable from a veridical case, yet the intermediate entities by which the subject accesses the world have misrepresented how things are. Since these entities are the subject’s only means of intercourse with the thought-independent world, their unreliability in cases of illusion raises the skeptical question of how the subject could determine that they do not regularly or comprehensively misrepresent the world. For given that they stop short of the world, once we allow that they sometimes misrepresent things, the worry arises that a wholesale gap could open up between how they represent the world and how the world in fact is, without the subject being able to discover this.
根据这种概念，认知错误 —— 错误信念 —— 实际上是主体对世界的表征与世界实际状态之间在准确性或对应性上的失败。幻觉是这种错误的一种，当主体的知觉经验与真实情况无法区分，但主体通达世界所凭借的中间实体却歪曲了事物的状态时，幻觉就会发生。由于这些实体是主体与独立于思想的世界交往的唯一途径，它们在幻觉案例中的不可靠性引发了一个怀疑论问题：主体如何能确定它们没有经常性或全面性地歪曲世界？因为鉴于它们与世界相隔绝，一旦我们承认它们有时会歪曲事物，就会产生这样的担忧：它们对世界的表征与世界的实际状态之间可能出现全面的鸿沟，而主体却无法发现这一点。

My suggestion, then, is that such skeptical worries are driven by a conception of the mind as engaging the world only indirectly, via a sort of representational intermediary.7 Without such an indirect, representational conception of mind, illusion is unlikely to generate skeptical anxiety. Indeed, absent such a conception of mind, philosophers would probably be more likely to dwell instead on the fact that in everyday life, we normally discover and correct for illusions as soon as they occur.8
因此，我的观点是，这种怀疑论担忧源于这样一种心灵概念：心灵只能通过某种表征性中介间接与世界打交道。7 没有这种间接的、表征性的心灵概念，幻觉就不太可能引发怀疑论焦虑。事实上，如果没有这样的心灵概念，哲学家们可能更倾向于关注这样一个事实：在日常生活中，我们通常会在幻觉一出现时就发现并纠正它们。8

8 A second important factor behind some versions of the argument from illusion may be a simple fallacy in quantifier logic: inferring from the premise that the senses deceive us in some cases to the conclusion that they might deceive us in all. I suggest, however, that a representational conception of mind helps explain why this faulty inference might seem persuasive.
8 某些版本的来自幻觉的论证背后的另一个重要因素可能是量词逻辑中的一个简单谬误：从感官在某些情况下欺骗我们这一前提出发，推出它们可能在所有情况下都欺骗我们这一结论。然而，我认为心灵的表征性概念有助于解释为什么这种错误推论可能看起来有说服力。

3 The Early Mohist Conception of Knowledge

3 早期墨家的知识概念

The Mòzǐ is the earliest Chinese text to reflect explicitly on knowledge and the grounds by which we obtain it. The Mohists understand knowledge in practical terms, mainly as a skill, ability, or competence. To know something is to be able to do something correctly-most fundamentally, to be able to draw distinctions properly.
《墨子》是中国最早明确反思知识及其获得依据的文献。墨家从实践角度理解知识，主要将其视为一种技能、能力或胜任力。认识某物就是能够正确地做某事 —— 最根本的是能够恰当地做出区分。

Mohist texts apply several interrelated conceptions of knowledge, or zhī 知. Probably the most common is a form of recognition, or “knowledge-of.”9 Knowledge-of is manifested as the practical ability to correctly discriminate the referent of a word, or “name” (míng 名), that denotes the object of knowledge. To qualify as having knowledge of oxen, for instance, the agent must be able to reliably pick out the sorts of things denoted by the word “ox.” The object of knowledge is typically regarded as a thing orevent, denoted by a term, rather than a fact or proposition, expressed by a sentence.10
墨家文献运用了几种相互关联的知识（知）概念。最常见的可能是一种识别，即 “关于…… 的知识”。9 “关于…… 的知识” 表现为能够正确辨别指称知识对象的词（“名”）的所指的实践能力。例如，要具备关于牛的知识，主体必须能够可靠地挑出 “牛” 这个词所指称的那类事物。知识的对象通常被视为由术语所指称的事物或事件，而非由句子所表达的事实或命题。10

9 Perhaps the most frequent use of the word zhī 知 (know) in early Chinese texts is in contexts in which it is best interpreted as “knowing-of” or “knowing-about,” a sort of recognition, familiarity, or understanding. A second common use is to mean roughly “know-to” or “know-how-to,” referring to a kind of competence or ability. Occasionally, zhī is used in contexts in which it is interpretable as “knowing that” and seems to refer to propositional knowledge. However, it is not clear that early Chinese thinkers themselves distinguish such contexts from “knowing-of.” In any case, the Mohists’ discussion of issues such as fatalism and the existence of ghosts indicates that the three sorts of knowledge are interrelated, and all three are explained by appeal to the ability to discriminate things properly: to know-that or know-of is to know-to discriminate different kinds of things correctly and apply the appropriate names or phrases to them. For opposing views, see Harbsmeier 1993 (also reprinted in Harbsmeier 1998), and Roetz 1993.
9 在中国早期文献中，“知”（know）一词最常见的用法可能是在最好被解释为 “关于…… 的知识” 或 “了解……” 的语境中，即一种识别、熟悉或理解。第二种常见用法大致意为 “知道去……” 或 “知道如何……”，指一种胜任力或能力。偶尔，“知” 也用于可被解释为 “知道……” 的语境中，似乎指命题性知识。然而，不清楚中国早期思想家自己是否将这类语境与 “关于…… 的知识” 区分开来。无论如何，墨家对宿命论和鬼神存在等问题的讨论表明，这三种知识是相互关联的，并且都通过恰当辨别事物的能力来解释：“知道……” 或 “关于…… 的知识” 就是知道正确区分不同种类的事物，并对它们使用恰当的名称或短语。关于相反的观点，参见哈布斯迈尔 1993（亦重印于哈布斯迈尔 1998）和罗茨 1993。

10 Grammatically, the classical Chinese verb zhī (to know) usually takes either a noun phrase or a verb phrase as its object. Even when the object of knowledge is a fact or state of affairs, it is normally expressed as a noun phrase, rather than a clause. For instance, instead of “How do we know that promoting the worthy is the root of government?” the Mohists write the Chinese equivalent of “How do we know promoting the worthy’s being the root of government?” (Mozi 9/2). Though I think it a mistake to read too much into contingent features of a thinker’s language, this grammar might have encouraged the Mohists to think of facts as features of things, structured along the lines of phrases predicable of things, rather than as states of affairs, with a propositional structure.
10 从语法上看，古汉语动词 “知”（to know）的宾语通常是名词短语或动词短语。即使知识的对象是事实或事态，它通常也被表达为名词短语，而非从句。例如，墨家不会写 “我们如何知道举贤是为政之本？”，而是写成相当于 “我们如何知道举贤之为政之本？”（《墨子》9/2）。尽管我认为过度解读思想家语言的偶然特征是错误的，但这种语法可能促使墨家将事实视为事物的特征，其结构类似于可归属于事物的短语，而非具有命题结构的事态。

This is not to say that the Mohists have no functional analogue to factual or propositional knowledge, or “knowledge-that.” The claim is rather that they conceptualize and explain such knowledge in terms of the ability to discriminate different kinds (lèi) of things. To know that a is F is to know to discriminate a as the kind of thing denoted by the term “F.” Also, since to know a thing is to know how to discriminate it from other kinds of things, the object of knowledge is sometimes taken to be a distinction, as when the Mohists speak of “knowing the distinction between moral and immoral” (Mozi 17/13; References to Mòzǐ cite chapter and line numbers in Mòzǐ 1986). What we think of as evaluating an assertion or determining the facts or truth about something for the Mohists is a process of drawing distinctions, such as discriminating shì 是 (this, right) from (fēi) 非 (not, wrong), (l i) 利 (benefit) from hài 害 (harm), yǒu 有 (presence) from wú 無 (absence), or tóng 同 (same) from yì 異 (different). For instance, in response to the fatalists’ claim that wealth, longevity, and other matters are all fixed by fate, Mòzǐ is depicted as asserting that “we cannot fail to clearly discriminate (biàn 辨)” the controversial doctrine (Mozi 35/5)- that is, we must evaluate whether it is shì or (fēi) and leads to benefit or harm.
这并不是说墨家没有与事实性知识或命题性知识（即 “知道……”）功能类似的概念。而是说，他们根据辨别不同种类（类）事物的能力来概念化和解释这类知识。知道 a 是 F，就是知道将 a 辨别为 “F” 这个术语所指称的那类事物。此外，由于认识某物就是知道如何将它与其他种类的事物区分开来，知识的对象有时被视为一种区分，例如墨家所说的 “知义与不义之别”（《墨子》17/13；《墨子》的引用标注依据《墨子引得》1986 年版的篇章和行号）。对墨家而言，我们所认为的对一个断言的评价，或对某物的事实或真相的确定，是一个做出区分的过程，例如区分是（此、对）与非（不、错）、利（益处）与害（害处）、有（存在）与无（不存在）、同（相同）与异（不同）。例如，针对宿命论者关于财富、长寿和其他事情都由命运决定的主张，墨子被描述为断言 “我们不能不明确辨别” 这一有争议的学说（《墨子》35/5）—— 也就是说，我们必须评估它是 “是” 还是 “非”，以及它会带来益处还是害处。

The Mohist conception of knowledge as an ability or competence is illustrated in the following passage, according to which knowledge lies not in making correct statements, but the ability to “select” the things denoted by a word, or “name.” The object of knowledge is not a proposition, but concrete objects-white and black items.
以下段落阐明了墨家将知识视为一种能力或胜任力的概念：知识不在于做出正确的陈述，而在于 “选取” 由词（“名”）所指称的事物的能力。知识的对象不是命题，而是具体的对象 —— 白色和黑色的物品。

Our master Mòzǐ said, “Now the blind say, ‘What’s bright is white, and what’s dark is black.’ Even the clear-sighted have no basis for changing this statement. But place white and black together and make the blind select among them, and they cannot know them. So as to my saying the blind do not know white and black, it is not on the basis of their naming; it is on the basis of their selecting.” (Mozi 47/23–24)
我们的墨子大师说：“现在盲人说：‘明亮的是白色，黑暗的是黑色。’即使是视力好的人也没有理由改变这种说法。但是把白色和黑色的东西放在一起，让盲人从中挑选，他们就无法识别它们。因此，我说盲人不知道白色和黑色，不是基于他们的命名，而是基于他们的挑选。”（《墨子》47/23–24）

The passage depicts the blind as possessing factual or propositional knowledge about the colors white and black. Yet it claims that the blind do not qualify as “knowing white and black,” because they are unable to discriminate white and black things in practice. The criterion of knowledge here is not understanding of a fact, nor the ability to state facts or use words correctly, but the ability to “distinguish the things” denoted by words (Mozi 19/5).11
这段话描述盲人拥有关于白色和黑色的事实性或命题性知识。然而，它主张盲人不配 “知道白色和黑色”，因为他们无法在实践中辨别白色和黑色的事物。这里的知识标准不是对事实的理解，也不是陈述事实或正确使用词语的能力，而是 “区分词语所指称的事物” 的能力（《墨子》19/5）。11

11 Against the Mohist position here, we might argue that this passage itself implicitly acknowledges that knowing how to use “names” is a discrete type of knowledge, distinct from knowing how to “select” things, and that the blind may possess the former even if they lack the latter. As we will see below, the later Mohist “Dialectics” incorporates this point into its account of knowledge.
11 针对墨家在这里的立场，我们可能会争辩说，这段话本身含蓄地承认，知道如何使用 “名” 是一种独立的知识类型，与知道如何 “挑选” 事物不同，盲人即使缺乏后者，也可能拥有前者。正如我们将在下面看到的，后期墨家 “辩经” 将这一点纳入了其知识理论中。

A further feature of the early Mohist conception of knowledge is that it includes no element corresponding to the justification condition in the traditional justified, true belief (“JTB”) account of knowledge familiar from the Western tradition. One reason the Mohists assign no role to justification is probably that their conception of knowledge as correct discrimination and response implicitly incorporates reliability. The main role of justification in the JTB account is to disqualify accidentally true beliefs, such as lucky guesses, from counting as knowledge. The Mohists implicitly handle this issue by considering only correct discrimination that issues from a reliable ability to be knowledge. Knowledge is not merely a matter of discriminating things correctly in one case or another, but a competence or reliable capacity to discriminate them correctly in a diverse range of cases (Mozi 17/11–12).
早期墨家知识概念的另一个特征是，它不包含与西方传统中熟悉的、传统的 “得到辩护的真信念”（JTB）知识理论中的辩护条件相对应的要素。墨家不赋予辩护任何角色的一个原因可能是，他们将知识视为正确的辨别和回应，这一概念含蓄地包含了可靠性。在 JTB 理论中，辩护的主要作用是将偶然为真的信念（如幸运的猜测）排除在知识之外。墨家通过只将源于可靠能力的正确辨别视为知识，含蓄地处理了这个问题。知识不仅仅是在某个或某些情况下正确辨别事物，而是在各种不同情况下正确辨别事物的一种胜任力或可靠能力（《墨子》17/11–12）。

For the Mohists, then, knowledge is not a matter of having mental states with certain features, such as being justified or true, but an ability to perform certain skills. Accordingly, cognitive error is not explained as a failure of the agent’s mental states to correspond to or represent the world accurately. It is understood as “disorder” (luàn 亂) or “confusion” (huò 惑) in discriminating things, in effect a failure to perform a skill correctly. For instance, the Mohists criticize those who fail to condemn unprovoked military aggression for the “disorder” in how they discriminate right from not-right (Mozi 17/14).
因此，对墨家而言，知识不在于拥有具有某些特征（如得到辩护或为真）的心理状态，而在于执行某些技能的能力。相应地，认知错误不被解释为主体的心理状态未能准确对应或表征世界，而是被理解为辨别事物时的 “乱”（混乱）或 “惑”（困惑），实际上是未能正确执行技能。例如，墨家批评那些不谴责无端军事侵略的人在辨别是与非时的 “混乱”（《墨子》17/14）。

How do we know we are discriminating such things correctly? To guide us in drawing and acting on distinctions properly, the Mohists introduce the concept of “models” or “standards” (fǎ 法), paradigms or criteria of the things we seek to identify or norms of conduct we aim to follow. They conceive of such models largely by analogy to measuring tools, exemplars, or benchmarks. Typical examples are artisan tools such as the carpenter’s setsquare or the wheelwright’s compass. Such models are applied to guide or check judgments or actions by evaluating whether the case at hand is relevantly similar to, and thus “coincides,” with them.
我们如何知道我们在正确地辨别这类事物？为了指导我们恰当地做出区分并依据区分行动，墨家引入了 “法”（模型或标准）的概念，即我们试图识别的事物的范例或标准，或我们旨在遵循的行为规范。他们主要通过与测量工具、范例或基准的类比来构想这些模型。典型的例子是工匠的工具，如木匠的矩尺或轮匠的圆规。应用这些模型来指导或检验判断或行动，方法是评估手头的案例是否与它们具有相关的相似性，从而 “符合” 它们。

The wheelwright and carpenter grasp their compass and setsquare and with them measure square and round in the world, saying, “What coincides is shì (this, right), what does not coincide is (fēi) (not-this, wrong).” (Mozi 26/41–42)
轮匠和木匠拿起他们的圆规和矩尺，用它们测量世界上的方形和圆形，说：“符合的就是‘是’（此、对），不符合的就是‘非’（非此、错）。”（《墨子》26/41–42）

Now the wheelwright takes his compass and uses it to measure the round and not-round in the world, saying, “What coincides with my compass, call it ‘round’; what doesn’t coincide with my compass, call it ‘not-round.’” Thus round and not-round can both be known. (27/63–65)
现在轮匠拿起他的圆规，用它测量世界上的圆形和非圆形，说：“与我的圆规相符合的，就称之为‘圆形’；与我的圆规不相符合的，就称之为‘非圆形’。” 因此，圆形和非圆形都可以被认识。（27/63–65）

The Mohists develop an explicit methodology for discriminating whether “statements” or “teachings” (yán 言) are shì or (fēi) by reference to three fundamental models: the historical precedent of the sage-kings; empirical evidence, specifically “the stuff of people’s ears and eyes” (what people hear and see);12 and beneficial practical consequences (Mozi 35/7–10). The issue of how we know something is thus typically resolved by citing one or more models and examining whether some claim or thing is sufficiently similar to and thus “coincides” with them.
墨家发展了一种明确的方法，通过参考三个基本模型来辨别 “言”（陈述或学说）的 “是” 与 “非”：圣王的历史先例；经验证据，特别是 “众人耳目之实”（人们听到和看到的东西）；12 以及有益的实际后果（《墨子》35/7–10）。因此，我们如何认识某物的问题，通常通过引用一个或多个模型，并考察某个主张或事物是否与它们足够相似，从而 “符合” 它们来解决。

12 The “stuff of people’s ears and eyes” has a double significance. On the one hand, it refers to empirical observation, by contrast with the authority of tradition. On the other, it refers to what the common people can determine for themselves, by contrast with what they accept on the authority of the political elite. I thank Jane Geaney for reminding me of the importance of social class in interpreting this criterion.
12 “众人耳目之实” 具有双重意义。一方面，与传统的权威相比，它指经验观察。另一方面，与人们基于政治精英的权威所接受的东西相比，它指普通民众能够自己确定的东西。感谢简・吉尼提醒我社会阶层在解释这一标准时的重要性。

4 Knowledge in the Mohist “Dialectics”

4 墨家 “辩经” 中的知识

As in earlier Mohist texts, epistemological passages in the later Mohist “Dialectics” focus not on the justification or truth of beliefs, or propositional attitudes, but on knowing how to distinguish what things are shì (this) or (fēi) (not) with respect to terms. The core expression of knowledge is again knowing how to apply terms to things correctly. However, the “Dialectics” identifies new aspects of knowledge to explain cases such as when an agent can correctly apply a term to a thing but not identify it perceptually or can recognize a thing under one name but not another. Here I will not present an exhaustive account of the epistemology of the “Dialectics,” but only survey a handful of passages relevant to perceptual knowledge.13
与早期墨家文献一样，后期墨家 “辩经” 中的认识论段落不关注信念或命题态度的辩护或真理性，而是关注知道如何依据术语区分事物的 “是”（此）与 “非”（非）。知识的核心表现再次是知道如何正确地将术语应用于事物。然而，“辩经” 识别了知识的新方面，以解释诸如主体能够正确地将一个术语应用于某物但无法通过知觉识别它，或者能够以一个名称识别某物但不能以另一个名称识别它等情况。在这里，我不会全面阐述 “辩经” 的认识论，而只是考察一些与知觉知识相关的段落。13

13 For a more detailed treatment, see Fraser 2009a. I should add the caveat that, as explained in Fraser 2009a, the textual problems with the Mohist “Dialectics” are so extensive and complex that any interpretation is partly conjectural. However, I think the key structural features emphasized here, such as the links between knowing, competence, discrimination, and language use, are among the aspects of later Mohist thought for which the textual evidence is quite strong. I should also point out that in calling a passage “relevant” to perceptual knowledge, I am not implying that it treats perceptual knowledge specifically and not other forms of knowledge as well.
13 更详细的论述参见弗雷泽 2009a。需要说明的是，正如弗雷泽 2009a 所解释的，墨家 “辩经” 的文本问题极为广泛和复杂，任何解释都在一定程度上是推测性的。然而，我认为这里强调的关键结构特征，如知识、胜任力、辨别和语言使用之间的联系，是后期墨家思想中 textual 证据相当充分的方面。我还应指出，称某一段落与知觉知识 “相关”，并不意味着它专门讨论知觉知识而不涉及其他形式的知识。

According to the later Mohists, humans who are alive and awake have a capacity or faculty called “the knowing” (zhī 知),14 roughly akin to the capacity to have conscious states and perform cognitive functions.15 The functions of “the knowing” overlap with those other early texts ascribed to the heart (xīn), a concept that plays no role in the “Dialectics.” One function of “the knowing” is to know (zhī 知) or recognize things in a manner the text describes as “contacting” them (A5), which probably refers to observation or perception.16 Characteristically, the Mohists explain this form of knowing by appeal to a practical ability: “To know is, by means of ‘the knowing,’ to pass by something and be able to characterize it.” So knowledge is the ability (néng 能) to characterize (mào 貌) something encountered, probably by applying one or more terms to it. It is not understood as a mental state or propositional attitude that represents the world accurately. Functionally, an instance of correctly applying a term to something-such as by saying, “This is an ox” or by pointing at an ox and declaring, “Ox!”-can be interpreted as an expression of propositional knowledge. But the text explains such knowledge in terms of an ability, most likely to distinguish and name things. As in early Mohist epistemology, justification plays no part in this conception of knowledge.
根据后期墨家的观点，活着且清醒的人有一种称为 “知”（认知能力）的能力或官能，14 大致类似于拥有意识状态和执行认知功能的能力。15 “知” 的功能与其他早期文献中归属于 “心” 的功能重叠，但 “心” 这一概念在 “辩经” 中没有作用。“知” 的功能之一是认识或识别事物，文本将这种方式描述为 “接触” 它们（A5），这可能指观察或知觉。16 典型地，墨家通过一种实践能力来解释这种认识形式：“知，接也。”（通过 “知” 这一能力，遇到某物并能够描述它）。因此，知识是描述（貌）所遇到的某物的能力（能），可能是通过对其应用一个或多个术语。它不被理解为准确表征世界的心理状态或命题态度。从功能上讲，正确地将一个术语应用于某物的例子 —— 例如说 “这是一头牛” 或指着一头牛宣称 “牛！”—— 可以被解释为命题性知识的表达。但文本根据一种能力来解释这种知识，很可能是区分和命名事物的能力。与早期墨家认识论一样，辩护在这种知识概念中没有任何地位。

14 See canon and explanation A3 and A22. References to the “Dialectics” follow the numbering system in Graham 2003.
14 参见经及经说 A3 和 A22。“辩经” 的引用遵循格雷厄姆 2003 年版的编号系统。

15 The faculty or capacity of zhī here can also be interpreted as “the awareness,” “consciousness,” or “the intelligence” (the latter is Graham’s translation [see Graham 2003: 267]). I render it “the knowing” partly to highlight the partial analogy to the Lockean or Kantian notion of “the understanding” and partly to emphasize that its primary function is to know things. I translate Xúnzǐ’s concept of zhēng zhī 徵知 (“verifying knowing”) similarly below.
15 这里的 “知” 这一官能或能力也可以被解释为 “意识”“觉知” 或 “智力”（后者是格雷厄姆的翻译 [见格雷厄姆 2003：267]）。我将其译为 “认知能力”，部分是为了强调它与洛克或康德的 “知性” 概念的部分类比，部分是为了强调其主要功能是认识事物。下面我对荀子的 “徵知”（“verifying knowing”）概念也采用类似的翻译。

16 I take canon A5 to refer to perceptual knowledge or, more precisely, knowledge originating in perception (which might include knowledge obtained from memory), because this seems the best explanation of the statement that knowledge is “contacting” (jiē 接) things. The text itself does not explicitly specify that A5 refers to perception.
16 我认为经 A5 指的是知觉知识，或更准确地说，源于知觉的知识（可能包括从记忆中获得的知识），因为这似乎是对知识是 “接触”（接）事物这一陈述的最佳解释。文本本身没有明确说明 A5 指的是知觉。

Another passage (A80) gives a categorization of what appear to be different sources and different forms or objects of knowledge.17 Among the sources are personal observation (qīn 親), which is explained as “observing it there oneself.” Along with the early Mohist notion of “the stuff of people’s ears and eyes,” this conception of “personal knowledge” (qīn zhì 親智，B70), or knowledge by observation, is the closest the Mohists come to articulating a notion of experience. Importantly, however, they themselves use no term actually interpretable as expressing a concept of experience. Nor do they posit epistemic or psychological intermediaries or representations between objects and the knowing agent, such as the ideas or impressions of Enlightenment conceptualism.18 Rather, perceived objects “enter” directly into “the knowing” (A98) by means of “the five routes,” or the sense organs (B46). “Personal knowledge” contrasts with “knowledge by explanation” (shuō zhì 說智，B70), which is in effect knowledge obtained by inference or reasoning, and knowledge by “hearsay” (wén 聞), which refers to the reports or testimony of others (A80, A81).
另一段落（A80）对似乎是知识的不同来源和不同形式或对象进行了分类。17 来源包括亲知（个人观察），被解释为 “身观焉”（亲自在那里观察）。连同早期墨家的 “众人耳目之实” 概念，这种 “亲智”（亲知）或通过观察获得的知识概念，是墨家最接近阐明经验概念的表述。然而，重要的是，他们自己没有使用任何实际上可被解释为表达经验概念的术语。他们也没有在对象和认识主体之间设定认知或心理中介或表征，如启蒙运动概念论中的观念或印象。18 相反，被感知的对象通过 “五路”（即感官）直接 “进入”“知”（A98）（B46）。“亲知” 与 “说智”（通过解释获得的知识）形成对比，后者实际上是通过推理获得的知识；还有 “闻知”（通过传闻获得的知识），指他人的报告或证词（A80，A81）。

17 Canon A80 lists seven items, of which the first three seem best explained as sources of knowledge: hearsay (wén 聞), explanation (shuō 說), and personal observation (qīn 親). The next four are names (míng 名), stuff (shí 實), matching (hé 合), and action (wéi 為). These seem best explained as either objects of knowledge or forms in which knowledge can be manifested-that is, by displaying the relevant competence in using names, discriminating stuff, matching names with stuff, or acting.
17 经 A80 列出了七项，其中前三项似乎最好被解释为知识的来源：闻知（传闻）、说知（解释）和亲知（个人观察）。接下来的四项是名（名称）、实（事物）、合（匹配）和为（行动）。这些似乎最好被解释为知识的对象或知识可以表现的形式 —— 即通过展示在使用名称、辨别事物、使名称与事物匹配或行动方面的相关胜任力。

18 As Geaney rightly observes, for early Chinese thinkers, “there is nothing to mediate and alienate the senses from the world” (Geaney 2002: 13).
18 正如吉尼正确观察到的，对中国早期思想家而言，“没有任何东西来中介感官与世界，使它们相疏离”（吉尼 2002：13）。

Among the different forms of knowledge (A80), with their corresponding objects, are knowledge of names (míng 名，words), of stuff (shí 實，including objects, events, and situations), and of the “matching” (hé 合) of names with stuff. Names are “that by which we call” (what we use to talk). Stuff is “what we call” (what we talk about). “Matching” is when name and stuff mate properly. The first two of these items concern what we might call “knowing-of.” Knowledge of names probably refers to knowing words, without necessarily knowing how to correctly discriminate what shí (stuff, things) they refer to. An example would be the blind people discussed earlier, who can correctly use the words “black” and “white” to state that “Bright things are white and dark things are black” but cannot identify black or white objects. Presumably the criterion for knowing a name is knowing how to use it properly in at least some contexts. Knowing “stuff” probably refers to recognizing objects, events, or situations under some name-at the very least, under the “all-reaching” name “thing” (wù 物，A78), which applies to everything-but without necessarily knowing the correct name for them. Knowledge of “matching” would be knowing how to correctly discriminate the kind of stuff or thing to which some name refers. This is the practical knowhow that the blind lack in the earlier example. This ability to match names with things may correspond to the “characterizing” of things discussed above (A5). Knowledge of names, stuff, and “matching” may all fall under the more general conception of discursive knowledge presented in canon A6, according to which knowledge is “understanding” (míng 明), manifested as an ability to use “the knowing” to “sort things” (lùn wù 論物)-that is, to distinguish them into kinds corresponding to various names or phrases.
在知识的不同形式（A80）及其相应对象中，包括对名（名称、词语）的知识、对实（事物，包括对象、事件和情境）的知识，以及对名与实的 “合”（匹配）的知识。名是 “所以谓”（我们用来称呼的东西）。实是 “所谓”（我们所称呼的东西）。“合” 是名与实恰当配合。前两项涉及我们可能称为 “关于…… 的知识” 的东西。对名的知识可能指知道词语，而不一定知道如何正确辨别它们所指的实（事物）。例如前面讨论的盲人，他们能正确使用 “黑” 和 “白” 这两个词来说 “明亮的是白色，黑暗的是黑色”，但不能识别黑色或白色的物体。大概知道一个名的标准是知道如何在至少某些语境中正确使用它。对 “实” 的知识可能指在某个名称下识别对象、事件或情境 —— 至少在适用于所有事物的 “达名”“物”（A78）之下 —— 但不一定知道它们的正确名称。对 “合” 的知识是知道如何正确辨别某个名称所指的那种事物。这是前面例子中盲人所缺乏的实践知识。这种使名与实匹配的能力可能对应于上面讨论的对事物的 “描述”（A5）。对名、实和 “合” 的知识可能都属于经 A6 提出的更一般的论述性知识概念，根据这一概念，知识是 “明”（理解），表现为运用 “认知能力” 来 “论物”（对事物进行分类）的能力 —— 即把它们区分为与各种名称或短语相对应的种类。

What determines whether a name and thing “match”? The Mohist explanation is again grounded in similarity to a model. A thing is “so” with respect to a predicate that is, a name and thing “match”-when its features are “similar” (ruò 若) to those of a model (fǎ 法) of the kind of thing in question (A70–71). Models may include yì 意，a thought or intention of the thing; measuring tools or other devices for identifying it; and exemplars of the thing. Canon A70, for instance, gives three examples of models, presumably for a circle or round object: a thought, a compass, and a circle. Features the Mohists cite as criteria of similarity include “shape and surface characteristics” (xíng mào 形貌) and “amount and number” (liàng shù 量數) (NO2). Things can “match” a model “exactly” (zhèng 正), as when a particular round figure is exactly the same shape as a model of a circle, or by whatever selected features are “appropriate” (yí 宜，A83), as when we deem people “swarthy” on the basis of the color of their skin, not their hair or eyes (A96). In some cases, discriminating different kinds can be difficult, because even if we have an exemplar of the kind to employ as a model, it may be hard to pin down just what features are “appropriate” or distinctive of that kind (B2).
什么决定名与实是否 “匹配”？墨家的解释再次基于与模型的相似性。当一个事物的特征与所讨论的那种事物的模型（法）的特征 “相似”（若）时，这个事物相对于一个谓词而言就是 “如此” 的 —— 即名与实 “匹配”（A70–71）。模型可能包括 “意”（对事物的想法或意图）、用于识别事物的测量工具或其他装置，以及事物的范例。例如，经 A70 给出了三个模型的例子，大概是针对圆形物体的：想法、圆规和一个圆形。墨家引用的作为相似性标准的特征包括 “形貌”（形状和表面特征）和 “量数”（数量）（NO2）。事物可以 “正”（精确地）与模型 “匹配”，例如某个圆形与圆形模型的形状完全相同；或者通过任何被 “宜”（适当）选择的特征，例如我们根据人的皮肤颜色而非头发或眼睛颜色来判定人是否 “黑”（A96）。在某些情况下，区分不同种类可能很困难，因为即使我们有该种类的一个范例作为模型，也可能难以确定哪些特征是该种类的 “适当” 或独特特征（B2）。

To discriminate kinds properly, then, an agent must know how to select and apply a suitable model, how to pick out the features of particular individuals that are the “appropriate” or relevant basis for evaluating similarity to or difference from the model, and how to discriminate whether those features indeed resemble the model.19 The agent must then successfully exercise these abilities by attending to the appropriate features, correctly judging whether they meet the threshold for similarity to the model, and disregarding other, irrelevant features. In discriminating whether someone is swarthy, for instance, he must know to disregard black hair and pupils and attend only to skin color. In discriminating horses from oxen, he must know not to attend to their having molars and a tail, since both kinds of animals have these features (B66).
因此，要恰当地辨别种类，主体必须知道如何选择和应用合适的模型，如何挑选特定个体的特征（这些特征是评估与模型相似或不同的 “适当” 或相关基础），以及如何辨别这些特征是否确实与模型相似。19 然后，主体必须通过关注适当的特征、正确判断这些特征是否达到与模型相似的阈值，并且忽略其他不相关的特征，成功地运用这些能力。例如，在辨别某人是否黑时，他必须知道忽略黑发和黑瞳孔，只关注皮肤颜色。在区分马和牛时，他必须知道不关注它们有臼齿和尾巴，因为这两种动物都有这些特征（B66）。

19 Though their texts do not explicitly say so, I assume the Mohists would agree that once we are thoroughly familiar with some kind, we can discriminate individuals of the kind without explicit reference to a model. The Mohists themselves are concerned mainly with cases in which confusion or disagreement might arise in discriminating kinds and so explicit comparison with models may be necessary.
19 尽管他们的文本没有明确说明，但我假设墨家会同意，一旦我们完全熟悉某一种类，我们就可以在不明确参考模型的情况下辨别该种类的个体。墨家自己主要关注的是在辨别种类时可能出现混淆或分歧的情况，因此可能需要与模型进行明确比较。

The “Dialectics” does not explicitly discuss the details of how agents make errors in discriminating different kinds of things. But given the above account of correct discrimination, we can make an informed conjecture as to how the later Mohists might explain perceptual error. The agent is likely to fail to discriminate correctly if he has not learned which features to attend to or how to gauge whether they count as similar to suitable models. He may also fail if he inadvertently neglects some relevant features, attends to irrelevant ones, or misjudges the degree of similarity to the model. Suppose the agent has indeed learned to discriminate individuals of the kind in question and can generally do so reliably, yet makes a mistake in a particular case. A likely cause is that the object he perceives is indeed similar to a relevant model, but only partly so. This partial similarity may lie in an inappropriate feature of the object, such as only its shape and not its surface features or, in the case of “swarthy,” a person’s hair instead of skin. Or it may lie in an appropriate feature, but be insufficient to qualify the object as an individual of the kind under consideration, as when the shape of an oval matches the shape of a model for “circle” partly but not entirely.
“辩经” 没有明确讨论主体在辨别不同种类事物时如何出错的细节。但根据上述对正确辨别的说明，我们可以合理推测后期墨家可能如何解释知觉错误。如果主体没有学会关注哪些特征，或者没有学会如何判断这些特征是否与合适的模型相似，他很可能无法正确辨别。如果他无意中忽略了一些相关特征、关注了不相关特征，或者错误判断了与模型的相似程度，也可能会失败。假设主体确实学会了辨别所讨论种类的个体，并且通常能可靠地做到这一点，但在特定情况下犯了错误。一个可能的原因是，他所感知的对象确实与相关模型相似，但只是部分相似。这种部分相似可能在于对象的一个不适当特征，例如只在于其形状而非表面特征，或者在 “黑” 的例子中，在于人的头发而非皮肤。或者，它可能在于一个适当的特征，但不足以使该对象成为所考虑种类的个体，例如椭圆形的形状与 “圆形” 模型的形状部分匹配但不完全匹配。

If this line of conjecture is correct, then we can understand why for the later Mohists perceptual error would not cast suspicion on the reliability of the senses, prompt the worry that appearances may fail to reflect reality, or raise the threat of losing touch with the world. Error occurs not when the world has somehow been misrepresented to the agent, but when the agent has discriminated on improper grounds-on the basis of partial but irrelevant or insufficient similarity between the object at hand and models of the kind under consideration. Not only does the agent remain in touch with the world, he has probably correctly noticed some similarity between the object and the kind.20 But he has overlooked other features pertinent to correct discrimination. In doing so, the agent has failed in the exercise of an ability at which he is normally competent. Thus understood, perceptual error results from faulty performance of a skill. It is comparable to a native speaker of a language absent-mindedly saying something grammatically incorrect. The locus of error is not the nature of perception, nor the agent’s relation to the world, but the agent’s inept performance: if the agent exercises his abilities properly, error will not occur.
如果这种推测正确，那么我们就能理解为什么对后期墨家而言，知觉错误不会让人怀疑感官的可靠性，不会引发表象可能无法反映实在的担忧，也不会带来与世界失去联系的威胁。错误的发生，不是因为世界以某种方式向主体呈现了歪曲的表象，而是因为主体基于不适当的理由进行了辨别 —— 基于手头对象与所考虑种类的模型之间部分但不相关或不充分的相似性。主体不仅与世界保持联系，他很可能还正确地注意到了对象与该种类之间的某些相似性。20 但他忽略了与正确辨别相关的其他特征。在这种情况下，主体在运用他通常能够胜任的能力时失败了。如此理解，知觉错误源于技能的不当执行。这类似于母语者心不在焉地说出语法错误的话。错误的根源不在于知觉的本质，也不在于主体与世界的关系，而在于主体的不熟练表现：如果主体恰当地运用其能力，错误就不会发生。

20 Let me emphasize that the point is not that the agent is correct about how things seem or appear. It is that he is correct about some feature of how they actually are.
20 我要强调的是，这一点并不是说主体对事物看起来或显现的样子是正确的，而是说他对事物实际所是的某些特征是正确的。

Another passage in the “Dialectics” identifies several sources of doubt or confusion and thus potential causes of erroneous discrimination (B10). The first of these is accidental or atypical circumstances, which the Mohists see as the cause of perceptual illusion, among other errors. One example of such circumstances is a dense fog, which might cause an agent to mistake a man for an ox, an instance of illusion. Another is unseasonable weather, which might cause a person who has moved to a summer hut, expecting a heat wave, to instead feel cold. In this case, the error is due not to illusion, but to the fallibility of induction. The second source of doubt is insufficient evidence. We have reason to doubt that someone who lifts a load of feathers is strong, since feathers are too light for lifting them to be evidence of strength. The third source is causal overdetermination. A fighter’s collapse might be because he drank alcohol before the match or because he was overcome by the noonday heat, and we may be unable to determine which factor was decisive. The fourth source is transience, when the object of discrimination is something that may cease. An example (the text does not give one) might be knowing the weather. When we fled indoors, we knew it was raining heavily, but the rain may have since stopped.

“辩经” 中的另一段落指出了怀疑或困惑的几个来源，因此也是错误辨别的潜在原因（B10）。第一个是偶然或非典型情况，墨家认为这是导致知觉幻觉和其他错误的原因。这种情况的一个例子是浓雾，它可能导致主体将人误认为牛，这是幻觉的一个实例。另一个例子是反常的天气，可能导致一个搬到避暑小屋、预期会有热浪的人反而感到寒冷。在这种情况下，错误不是由于幻觉，而是由于归纳的可错性。怀疑的第二个来源是证据不足。我们有理由怀疑一个举起一担羽毛的人是否强壮，因为羽毛太轻，举起它们不能作为强壮的证据。第三个来源是因果决定过多。一个战士的倒下可能是因为他在比赛前喝了酒，或者因为被正午的炎热所征服，而我们可能无法确定哪个因素是决定性的。第四个来源是暂时性，即被辨别的对象是可能消失的东西。一个例子（文本没有给出）可能是对天气的了解。当我们逃进室内时，我们知道正在下大雨，但那之后雨可能已经停了。

For our present purposes, the chief significance of this passage is that the fallibility of sense perception is not identified as grounds for confusion or error. Indeed, illusion or faulty perception is not even considered an independent category of error, but only one type of confusion due to unusual circumstances. The sources of confusion and potential error that concern the Mohists lie not in the nature of perception, nor the agent’s relation to the world, but in contingent, well-understood features of the agent s circumstances- that things are unclear, atypical, causally overdetermined, transient, or provide inadequate evidence. With the possible exception of causal overdetermination, in each case confusion or doubt can usually be resolved by further investigation.
就我们当前的目的而言，这段话的主要意义在于，感官知觉的可错性并未被视为困惑或错误的根源。事实上，幻觉或错误的知觉甚至不被视为一个独立的错误类别，而只是由于非典型情况导致的一种困惑。墨家所关注的困惑和潜在错误的来源，不在于知觉的本质，也不在于主体与世界的关系，而在于主体处境中偶然的、易于理解的特征 —— 事物不清晰、非典型、因果决定过多、具有暂时性，或者提供的证据不足。除了因果决定过多的情况可能例外，在其他每种情况下，困惑或怀疑通常都可以通过进一步的调查来解决。

5 Xúnzǐ on Knowledge

5 荀子论知识

Xúnzǐ follows the Mohists in tying knowledge to the ability to discriminate and name things. Like the later Mohists, he refers to the capacity or faculty of knowledge as “the knowing” (zhī 知) (Xúnzǐ 22/5).21 Knowledge is discriminating shì from fēi properly (Xúnzǐ 2/12), such that the attitudes of “the knowing” “match” (hé 合) the distinctions between things (Xúnzǐ 22/5). Distinctions underlie the use of names. The purpose of names is to discriminate different social ranks and similar from different things, so that intentions can be conveyed and tasks carried out (Xúnzǐ 22/14–15). Names can be used this way because we are all creatures of the same kind, with the same sort of dispositions (qíng 情). Thus our sense organs detect things similarly, enabling us to establish conventions for the use of names to discriminate different kinds of things (Xúnzǐ 22/16–17).22 Similarities and differences between features of things are “differentiated” (yì 異) by means of the sense organs—the eyes, ears, mouth, nose, and body (by which we differentiate pain, itching, temperature, texture, and weight). The heart (xīn) differentiates cognitive, affective, and conative features.
荀子与墨家一样，将知识与辨别和命名事物的能力联系在一起。与后期墨家类似，他将知识的能力或官能称为 “知”（认知能力）（《荀子》22/5）。21 知识是对是与非的恰当辨别（《荀子》2/12），使得 “知” 的态度与事物之间的区分 “合”（匹配）（《荀子》22/5）。区分是名称使用的基础。名称的目的是辨别不同的社会等级以及相同与不同的事物，以便能够传达意图和执行任务（《荀子》22/14–15）。名称能够以这种方式使用，因为我们都是同类生物，具有相同的性情（情）。因此，我们的感官对事物的感知是相似的，这使我们能够建立名称使用的惯例，以辨别不同种类的事物（《荀子》22/16–17）。22 事物特征之间的同与异通过感官 —— 眼睛、耳朵、嘴巴、鼻子和身体（我们通过身体区分痛、痒、温度、质地和重量）来 “异”（区分）。心区分认知、情感和意动特征。

21 References to the Xúnzǐ give chapter and line numbers in Xúnzǐ 1966.
21 《荀子》的引用标注依据《荀子引得》1966 年版的篇章和行号。

22 I follow Lǐ and Hansen (see Lǐ 1979: 513 and Hansen 1992: 325) in reading “all those of the same kind and same dispositions” as the antecedent of the pronoun qí 其 (their) in the phrase qí tiān guān 其天官 (their sense organs) at Xúnzǐ 22/16, for I see no other plausible antecedent. The verb phrase yì wù 意物 at Xúnzǐ 22/16 I take to be roughly equivalent to yì wù 臆物 (to detect or measure things), yì 意 being used in a way similar to how it is at Xúnzǐ 26/14. A crucial interpretive point is that yì wù is unlikely to refer to mental representations or images of things, as, for instance, Knoblock and Graham interpret it (Knoblock 1994: 129, and Graham 2003: 213). In contexts where it is used as a verb, as it is here, yì 意 can be interpreted as “intend,” “think,” “detect,” “assess,” or “suspect,” but is unlikely to mean “represent” or “form a mental image,” as it refers to the directedness of the sense organs toward their object, rather than an “input” from the object through the senses to the heart. Moreover, were the text referring to images or representations, it would likely use the word xiàng 象，as it does a few lines later when it states that the heart “represents” (xiàng) the dào through verbal discrimination and explanation (Xúnzǐ 22/40). Geaney correctly rejects the “mental representation” interpretation but takes yì wù to be a noun compound meaning roughly “intentional thing” (Geaney 2002: 37–39). This reading is difficult to defend grammatically, since it makes the nominalizing particle yě 也 redundant.
22 我遵循李和汉森（见李 1979：513 和汉森 1992：325）的观点，将 “凡同类同情者” 视为《荀子》22/16 中 “其天官”（他们的感官）一词中代词 “其”（他们的）的先行词，因为我没有看到其他合理的先行词。“意物” 这一动词短语在《荀子》22/16 中，我认为大致相当于 “臆物”（察觉或衡量事物），“意” 的用法类似于《荀子》26/14 中的用法。一个关键的解释要点是，“意物” 不太可能指事物的心理表征或意象，例如诺布洛克和格雷厄姆对其的解释（诺布洛克 1994：129，格雷厄姆 2003：213）。在像这里这样 “意” 用作动词的语境中，它可以被解释为 “意图”“思考”“察觉”“评估” 或 “怀疑”，但不太可能意为 “表征” 或 “形成心理意象”，因为它指的是感官指向其对象的指向性，而非从对象通过感官到心的 “输入”。此外，如果文本指的是意象或表征，它很可能会使用 “象” 这个词，就像几行之后所说的，心通过言语辨别和解释来 “象”（表征）道（《荀子》22/40）。吉尼正确地拒绝了 “心理表征” 的解释，但将 “意物” 视为一个名词复合词，大致意为 “意向性事物”（吉尼 2002：37–39）。这种解读在语法上难以成立，因为它使名词化助词 “也” 变得多余。

The heart also has a function called “the verifying knowing” (zhēng zhī 徵知), which is responsible for cognition (Xúnzǐ 22/19). The “verifying knowing” is able to recognize sounds by means of the ears, shapes by means of the eyes, and so on. The sense organs “record” (bù 簿) the features of things—sounds, shapes, and so on—and the heart “verifies,” or recognizes, them (Xúnzǐ 22/20). For an agent to qualify as having perceptual knowledge of something, her sense organs must “record” and become aware of it and the heart must “verify” it, such that the agent is able to “explain” (shuō 說) it (22/20–21). The text does not specify exactly what is meant by “explain” here. Probably the requirement is that the agent be able to correctly apply a name to the thing, and perhaps also to explain the grounds for discriminating the thing as belonging to the kind denoted by that name. If so, then this requirement is similar to that in the “Dialectics” that to have knowledge of something, an agent must be able to “characterize” it (A5) or “sort” it (A6). In both cases, knowing how to apply a word to a thing is a criterion of knowledge, which is inherently linguistic or conceptual.23
心还有一种称为 “徵知”（验证性认知）的功能，负责认知（《荀子》22/19）。“徵知” 能够通过耳朵识别声音，通过眼睛识别形状，等等。感官 “簿”（记录）事物的特征 —— 声音、形状等等 —— 而心 “验证” 或识别它们（《荀子》22/20）。一个主体要具备对某物的知觉知识，她的感官必须 “记录” 并意识到该事物，心必须 “验证” 它，使得主体能够 “说”（解释）它（22/20–21）。文本没有确切说明这里的 “说” 是什么意思。可能的要求是，主体能够正确地将一个名称应用于该事物，或许还能解释将该事物辨别为那个名称所指称的种类的依据。如果是这样，那么这一要求类似于 “辩经” 中的要求，即要拥有对某物的知识，主体必须能够 “描述” 它（A5）或 “分类” 它（A6）。在这两种情况下，知道如何将一个词应用于一个事物都是知识的一个标准，而知识本质上是语言的或概念的。23

23 Geaney suggests that since the text places a pragmatic condition on knowing, the form of knowledge it articulates is not perception (Geaney 2002: 48–49, 203 fn. 85). Her reading is intriguing, but I think a simpler yet equally comprehensive interpretation is that the text is articulating a conception of perceptual knowledge or recognition that incorporates a pragmatic requirement. This interpretation coheres with Xúnzǐ’s view that the activity of the sages is what divides things in the natural world into kinds (Xúnzǐ 19/ 78–79) and thus underwrites our ability to have knowledge of those kinds. One must grasp the traditional, conventional distinctions embodied in language, as passed down from the sage-kings, in order to qualify as having perceptual knowledge.
23 吉尼认为，由于文本对知识设定了一个实用条件，它所阐述的知识形式不是知觉（吉尼 2002：48–49，203 注 85）。她的解读很有趣，但我认为一种更简单但同样全面的解释是，文本正在阐述一种包含实用要求的知觉知识或识别概念。这种解释与荀子的观点一致，即圣人的活动将自然界中的事物划分为不同种类（《荀子》19/78–79），从而支持我们拥有关于这些种类的知识的能力。一个人必须掌握从圣王那里传下来的、体现在语言中的传统的、约定俗成的区分，才能有资格拥有知觉知识。

A further similarity to the “Dialectics” is that Xúnzǐ’s theory of perceptual knowledge assigns no role to anything comparable to the Cartesian-Empiricist notion of ideas or impressions, nor to any other form of epistemic or psychological intermediary.24 The text’s theoretical framework comprises only features of things such as shape, sound, sweetness, odor, or heat—the sense organs, which differentiate these features, and the heart, which employs the sense organs to “verify” or recognize them.
与 “辩经” 的另一个相似之处是，荀子的知觉知识理论没有给任何可与笛卡尔 - 经验主义者的观念或印象概念相比拟的东西，或任何其他形式的认知或心理中介分配角色。24 文本的理论框架仅包括事物的特征，如形状、声音、甜味、气味或热度 —— 区分这些特征的感官，以及运用感官来 “验证” 或识别它们的心。

24 Here I agree wholly with Geaney in rejecting interpretations (such as those in Cua 1985: 31, Watson 1963: 142–43, and Knoblock 1994: 130) that ascribe a role to sense data or sense impressions in Xúnzǐ’s theory (Geaney 2002: 201, n. 83). The text explicitly discusses only the “differentiating” activity of the sense organs and heart and the “verifying” activity of the heart (Xúnzǐ 22/16–21). It mentions nothing corresponding to sense impressions or sense data.
24 在这里，我完全同意吉尼的观点，拒绝那些认为荀子的理论中感觉材料或感官印象有一定作用的解释（如匡 1985：31，沃森 1963：142–43，诺布洛克 1994：130）（吉尼 2002：201，注 83）。文本明确只讨论了感官和心的 “区分” 活动以及心的 “验证” 活动（《荀子》22/16–21）。它没有提到任何与感官印象或感觉材料相对应的东西。

6 Xúnzǐ on Cognitive Error

6 荀子论认知错误

An entire book of the Xúnzǐ—“Resolving Obscuration”—is devoted to the topic of cognitive error. This text explicitly presents what I call the early Chinese “part-whole” conception of knowledge and error. The book’s main theme is that errors in thought and action—errors in discriminating things and responding to them correctly—are due to what it calls “obscuration” or “blinkering (bì 蔽),” which can be avoided by employing the heart properly. Error occurs when agents become fixated on, and thus “obscured” or “blinkered” by, one “bend (qū 曲)” or “corner (yú 隅)” of things, leaving them ignorant of the “overall pattern (dà lǐ 大理)” (Xúnzǐ 21/1, 21/25). It is due not to subjective misrepresentation of the mind-independent world, nor to a gap between appearance and reality, but to fixing our attention in the wrong direction, such that we consider only some factors rather than all those relevant to discrimination. Erroneous discrimination is thus not wholly mistaken. The problem is that it is biased or one-sided and so only partly correct.25
《荀子》中有一整篇 ——《解蔽》—— 专门讨论认知错误的主题。这篇文本明确提出了我所谓的中国早期知识与错误的 “部分 - 整体” 概念。该书的主要主题是，思想和行动中的错误 —— 在辨别事物和对其做出正确回应时的错误 —— 是由于所谓的 “蔽”（遮蔽或蒙蔽），这可以通过正确运用心来避免。当主体专注于事物的一个 “曲”（方面）或 “隅”（角落），从而被其 “蔽”（遮蔽或蒙蔽），而对 “大理”（整体模式）一无所知时，错误就会发生（《荀子》21/1，21/25）。错误的产生，不是由于对独立于心灵的世界的主观歪曲，也不是由于表象与实在之间的鸿沟，而是由于我们将注意力固定在错误的方向上，以至于我们只考虑了某些因素，而不是所有与辨别相关的因素。错误的辨别因此并非完全错误。问题在于它带有偏见或片面性，因此只部分正确。25

25 Dan Robins also notes these aspects of Xúnzǐ’s explanation of error (Robins 2008: Section 8).
25 丹・罗宾斯也注意到了荀子对错误的解释的这些方面（罗宾斯 2008：第 8 节）。

According to Xúnzǐ, such bias occurs because the heart is disturbed, misdirected, or hampered by internal conditions, such as emotions or obsessions, or external conditions, such as darkness, interference with the sense organs, or alcohol. These conditions impede the agent in responding to relevant aspects of the “overall pattern” in some situations. Conscientious, competent agents can avoid error, because they know how to employ the heart so as to prevent such conditions from interfering with correct discrimination. Perceptual error thus is not ascribed to flaws in the functioning of the sense organs, nor is it considered evidence that perception is inherently deceptive or unreliable. It occurs not because perception or the senses are untrustworthy, but because agents have failed to employ their capacities properly.26
根据荀子的观点，这种偏见的产生是因为心受到内部条件（如情绪或执念）或外部条件（如黑暗、感官受到干扰或酒精）的干扰、误导或阻碍。在某些情况下，这些条件阻碍主体对 “大理” 的相关方面做出回应。认真、有能力的主体可以避免错误，因为他们知道如何运用心，以防止这些条件干扰正确的辨别。因此，知觉错误不被归因于感官功能的缺陷，也不被视为知觉本质上具有欺骗性或不可靠的证据。它的发生不是因为知觉或感官不可信，而是因为主体未能正确运用其能力。26

Xúnzǐ applies this theory to explain error on two levels, the general level of recognizing the overall dào 道 (way, teachings) or values one should live by and the concrete level of conducting oneself according to a particular dào. As examples of error on the general level, he criticizes other philosophers who emphasize some values but, according to Xúnzǐ, are blinkered to others. The problem with rival thinkers such as Mòzǐ or Zhuāngzǐ 莊子，Xúnzǐ thinks, is not that their teachings are utterly wrong, but that they are partial, one-sided, or incomplete. Each possesses only “one corner of the dào” and is blind to the rest (Xunzi 21/24–25). Such blinkering typically occurs when people are partial to what they have already learned and prefer not to hear of its flaws or the advantages of other approaches (Xunzi 21/3–4). It thus prevents them from seeking what is right (Xunzi 21/4).
荀子将这一理论应用于解释两个层面的错误：认识人应遵循的整体 “道”（方式、教义）或价值的一般层面，以及根据特定 “道” 行事的具体层面。作为一般层面错误的例子，他批评了其他一些哲学家，这些哲学家强调某些价值，但在荀子看来，他们对其他价值视而不见。荀子认为，像墨子或庄子这样的对立思想家的问题，不在于他们的学说完全错误，而在于它们是片面的、单方面的或不完整的。每个人都只拥有 “道之一隅”，而对其余部分视而不见（《荀子》21/24–25）。这种蒙蔽通常发生在人们偏爱自己已经学到的东西，不愿意听到其缺陷或其他方法的优点时（《荀子》21/3–4）。因此，这阻碍了他们寻求正确的东西（《荀子》21/4）。

Almost anything can cause one-sidedness or blinkering, according to Xúnzǐ. His fundamental explanation of the cause of blinkering is that the various differences between things tend to obscure each other, thus disrupting our ability to “sort” or “grade” (lún 倫) things properly (Xunzi 21/7, 21/29). (Recall that the ability to sort things correctly is also a criterion of knowledge for the Mohist “Dialectics,” according to the conception of “understanding” presented in canon A6.) The text does not elaborate, but Xúnzǐ’s idea is probably that one-sidedly or injudiciously attending to any one distinction or value makes us prone to overlook others that may also be pertinent. This failing is the crux of his attack on the one-sidedness of rival thinkers: he claims that they each emphasize one value or base their dào on one notion or technique at the expense of other equally important ones (Xúnzǐ 21/21–26). Mòzǐ, for instance, was “blinkered by utility and did not know cultural form” (Xúnzǐ 21/21): he focused so one-sidedly on material utility—and thus the benefit/harm distinction—that he neglected the importance of cultural or aesthetic form. In such cases, says Xúnzǐ, our recognition of “that one”—one side of a distinction—interferes with our understanding of “this one”—whatever falls on the other side (Xúnzǐ 21/38). A further dimension of the problem is that any distinguishing feature we attend to in discriminating things biases us toward understanding them specifically by reference to that feature, when other patterns of similarity and difference are always present as well. A trivial example is that in attending to what is desirable about something, we may overlook what is detestable about it, and in attending to what is beneficial about it, we may overlook what is harmful (Xúnzǐ 3/45–49). On a deeper level, any values or norms we live by will tend to blinker us toward alternative values and ways of life.27
根据荀子的观点，几乎任何事物都可能导致片面性或蒙蔽。他对蒙蔽原因的根本解释是，事物之间的各种差异往往相互遮蔽，从而破坏我们 “伦”（分类或分级）事物的能力（《荀子》21/7，21/29）。（回想一下，根据经 A6 提出的 “理解” 概念，正确分类事物的能力也是墨家 “辩经” 中知识的一个标准。）文本没有详细阐述，但荀子的想法可能是，片面地或不明智地关注任何一个区分或价值，会使我们容易忽视其他可能也相关的区分或价值。这一缺陷是他攻击对立思想家片面性的核心：他声称他们每个人都强调一种价值，或将其道建立在一个概念或技术之上，而牺牲了其他同样重要的东西（《荀子》21/21–26）。例如，墨子 “蔽于用而不知文”（《荀子》21/21）：他如此片面地关注物质效用 —— 从而关注利害区分 —— 以至于忽视了文化或审美形式的重要性。荀子说，在这种情况下，我们对 “彼”（一个区分的一面）的认识，会干扰我们对 “此”（属于另一面的任何东西）的理解（《荀子》21/38）。这个问题的另一个层面是，在辨别事物时，我们所关注的任何显著特征都会使我们倾向于专门通过该特征来理解它们，而其他相似性和差异性模式也总是存在的。一个微不足道的例子是，在关注某物的可取之处时，我们可能会忽视它的可憎之处；在关注它的有益之处时，我们可能会忽视它的有害之处（《荀子》3/45–49）。在更深层次上，我们所遵循的任何价值或规范，都会倾向于使我们对替代价值和生活方式视而不见。27

26 Geaney makes a series of similar observations (Geaney 2002: 40–41).
26 吉尼做出了一系列类似的观察（吉尼 2002：40–41）。

27 As we have seen, for both Xúnzǐ and the Mohists, cognition is based on discriminating different kinds of things. So Xúnzǐ’s account implies that some form of blinkering is inherent in the very structure of cognition and that any values we hold or dào we follow in some sense blinker us to alternative values and dào. This point is among the grounds for both the skeptical arguments and the normative stance of the Zhuāngzǐ. For further discussion, see Fraser 2006, Fraser 2009. Ironically, Xúnzǐ himself does not note these consequences of his view, but instead asserts that the dào of Confucius is comprehensive and unblinkered (Xúnzǐ 21/27).
27 正如我们所看到的，对荀子和墨家而言，认知都基于辨别不同种类的事物。因此，荀子的解释意味着某种形式的蒙蔽是认知结构本身所固有的，并且我们所持有的任何价值或所遵循的任何道，在某种意义上都会使我们对替代价值和道视而不见。这一点是《庄子》的怀疑论论证和规范性立场的依据之一。更多讨论参见弗雷泽 2006，弗雷泽 2009。具有讽刺意味的是，荀子本人没有注意到他的观点的这些后果，而是断言孔子的道是全面的、不受蒙蔽的（《荀子》21/27）。

The second level of error Xúnzǐ considers is performance error in following a particular dào. Even if an agent is committed to following the right norms or dào, the agent may still commit errors in concrete situations. I will discuss this form of error in the section below on Xúnzǐ’s treatment of illusion, where I will argue that his underlying explanation again lies in one-sidedness. Error occurs when the agent has attended to some features of the situation while overlooking other relevant features.
荀子考虑的第二个层面的错误是遵循特定 “道” 时的执行错误。即使主体致力于遵循正确的规范或 “道”，在具体情况下仍可能犯错。我将在下面关于荀子对幻觉的处理的部分讨论这种错误形式，我会论证他的根本解释仍然在于片面性。当主体关注了情境的某些特征而忽略了其他相关特征时，错误就会发生。

Xúnzǐ’s part-whole conception of error and knowledge articulates ideas that were apparently widely shared among third-century B.C.E. Chinese intellectuals. Like Xúnzǐ, several books of the Zhuāngzǐ criticize proponents of other views for their “partial” grasp of the dào.28 Two chapters of the The Annals of Lü Buwei 呂氏春秋 advance a part-whole explanation of error similar to Xúnzǐ’s.29 Rather than using Xúnzǐ’s term bì, the Annals ascribes error or misjudgment to being “biased” or “confined” (yòu 囿), typically by personal preferences. “When people are confined by something, indeed they will take daylight to be darkness, white to be black, and a sage-king like Yáo to be a tyrant like Jié. Being confined makes for serious mistakes” (Knoblock and Riegel 2000: 400).30 Like Xúnzǐ, the text attributes such bias to how agents direct their attention: “One who looks toward the east does not see the west wall; one who gazes toward the south does not observe the north. The intention has something it is fixed on” (Knoblock and Riegel 2000: 286). Xúnzǐ’s part-whole account also coheres well with the explanation of error that I suggested can be derived from the Mohist “Dialectics.” On my interpretation, for the Mohists, error occurs when the agent attends to the wrong features of things or overlooks relevant features. Xúnzǐ would explain this by saying that the agent has been blinkered, either by incorrect training or by attending to some features at the expense of others.
荀子的错误与知识的部分 - 整体概念所阐述的思想，显然在公元前三世纪的中国知识分子中得到了广泛认同。与荀子一样，《庄子》的几篇文章批评其他观点的支持者对 “道” 的 “片面” 把握。28 《吕氏春秋》的两章提出了与荀子类似的对错误的部分 - 整体解释。29 与荀子使用 “蔽” 一词不同，《吕氏春秋》将错误或误判归因于 “囿”（受偏见或局限），通常是由个人偏好导致的。“人固囿于其所已知，而未知其所未知也。夫道者，体常而尽变，一隅不足以举之。曲知之人，观于道之一隅而未之能识也。故以为足而饰之，内以乱其身，外以惑其人，上以蔽其君，下以惑其民，此蔽塞之祸也。”（诺布洛克和里格尔 2000：400）。30 与荀子一样，文本将这种偏见归因于主体如何引导他们的注意力：“东面而望，不见西墙；南面而视，不睹北方。意有所在也。”（诺布洛克和里格尔 2000：286）。荀子的部分 - 整体解释也与我提出的可从墨家 “辩经” 中推导出来的错误解释高度一致。根据我的解释，对墨家而言，当主体关注事物的错误特征或忽略相关特征时，错误就会发生。荀子会解释说，这是因为主体受到了蒙蔽，要么是由于不正确的训练，要么是因为关注某些特征而牺牲了其他特征。

28 See, for example, Zhuāngzǐ 13/40, 17/6ff., 24/33ff., 25/74, and 33/11ff. (The last passage is probably a Hàn dynasty text.) (Citations to the Zhuāngzǐ give chapter and line numbers in Zhuāngzǐ 1956.) A. C. Graham long ago noted that criticism of opponents’ views as narrow or one-sided rather than outright wrong is common in third century B.C.E. and later texts (Graham 1989: 378). Much of the discussion in Book 2 of the Zhuāngzǐ is driven by the thought that in discriminating things one way rather than another, there is always that which we overlook and thus fail to see (Zhuāngzǐ 2/57–58).
28 例如，见《庄子》13/40，17/6ff.，24/33ff.，25/74 和 33/11ff.（最后一段可能是汉代文本）。（《庄子》的引用标注依据《庄子引得》1956 年版的篇章和行号。）A.C. 格雷厄姆很久以前就指出，在公元前三世纪及以后的文本中，批评对手的观点是狭隘的或片面的，而非完全错误的，这是常见的（格雷厄姆 1989：378）。《庄子》第二篇中的许多讨论都是由这样一种想法推动的，即在以一种方式而非另一种方式辨别事物时，总有我们所忽视因而未能看到的东西（《庄子》2/57–58）。

29 Compare Knoblock and Riegel 2000: Sect. 13.3 and 16.7. Translations from the text are my own.
29 比较诺布洛克和里格尔 2000：第 13.3 节和第 16.7 节。文本的翻译是我自己的。

30 The Annals uses the graphs 宥 and 尤 for the word yòu instead of 囿 (see Zhuāngzǐ 24/34, 33/80), but all three graphs seem to express the same word.
30 《吕氏春秋》用 “宥” 和 “尤” 来表示 “yòu” 这个词，而不是 “囿”（见《庄子》24/34，33/80），但这三个字形似乎表达的是同一个词。

6.1 Avoiding Error

6.1 避免错误

This and the following subsections review in detail Xúnzǐ’s discussion of the causes of error and how to avoid or resolve them. The purpose of these sections is to illustrate how the details of Xúnzǐ’s theory cohere with a competence model of knowledge and a part-whole model of error.
本节及以下小节详细回顾荀子对错误原因以及如何避免或解决错误的讨论。这些部分的目的是说明荀子理论的细节如何与知识的胜任力模型和错误的部分 - 整体模型相一致。

Xúnzǐ regards the use of the heart to discriminate things and guide action as a field of skill or technique, which he calls “arts of the heart (xīn shù 心術)” (Xúnzǐ 21/28). As with any skill or art, performance in discrimination can be improved through training and conscientiousness. For Xúnzǐ, as for many early Chinese thinkers (excepting the Mohists and the Zhuāngzǐ, which parodies the idea), the heart is the “ruler (jūn 君)” of the other organs (Xúnzǐ 21/44, 17/12).31 As we have seen, the heart is responsible for discriminating things on the basis of differences between them detected via the sense organs. Through its discriminating function, the heart handles cognition, judgment, and reasoning. But Xúnzǐ sees these as subsidiary operations; the heart’s chief function is to guide action.32 In one place, Xúnzǐ describes the heart as the “labor supervisor (gōng zǎi 工宰)” (Xúnzǐ 22/40) of the dào—it is responsible for managing the other organs and the person as a whole so that the work of following the dào is performed properly. For the agent to discriminate and act according to the dào, the heart must be employed effectively. On their own, the sense organs can be obscured (bì) by external conditions, as when darkness obscures vision. But such obscuration need not lead to erroneous discrimination if the heart is employed competently. On the other hand, “if the heart is not employed at it,” Xúnzǐ says, “though white and black are before us, the eyes do not see them; though thunder drums are beside us, the ear does not hear them” (Xúnzǐ 21/4–5).
荀子将运用心来辨别事物和指导行动视为一种技能或技艺领域，他称之为 “心术”（心的技艺）（《荀子》21/28）。与任何技能或技艺一样，辨别能力可以通过训练和认真态度来提高。与许多中国早期思想家一样（除了墨家以及《庄子》——《庄子》对这一观点进行了调侃），荀子认为心是其他器官的 “君”（统治者）（《荀子》21/44，17/12）。31 正如我们所看到的，心负责根据通过感官检测到的事物之间的差异来辨别事物。通过其辨别功能，心处理认知、判断和推理。但荀子认为这些是次要操作；心的主要功能是指导行动。32 在某处，荀子将心描述为道的 “工宰”（工头）（《荀子》22/40）—— 它负责管理其他器官和整个人，以便正确执行遵循道的工作。为了让主体根据道进行辨别和行动，必须有效地运用心。感官本身可能会被外部条件所 “蔽”（遮蔽），例如黑暗遮蔽视觉。但如果心被胜任地运用，这种遮蔽不一定会导致错误的辨别。另一方面，荀子说：“心不使焉，则白黑在前而目不见，雷鼓在侧而耳不闻”（《荀子》21/4–5）。

31 For discussions of the early Chinese metaphor of the heart as political sovereign and the sense organs as bureaucrats, see Sivin 1995 and Geaney 2002: 17–22.
31 关于中国早期将心比作君主、将感官比作官僚的隐喻的讨论，见席文 1995 和吉尼 2002：17–22。

32 Donald Munro was the first to call attention to this aspect of the role of the heart in early Chinese thought and the intriguing contrast between it and Greek conceptions of the mind, which instead emphasize its role in knowledge and reasoning 1969: 51–59.
32 唐纳德・芒罗是第一个注意到中国早期思想中心的这一作用方面的人，以及它与希腊心灵概念之间的有趣对比 —— 希腊的心灵概念强调心灵在知识和推理中的作用（1969：51–59）。

To discriminate properly, the heart must maintain a balanced, attentive equilibrium, like that of a soloist awaiting the conductor’s cue or a basketball player compensating for an opponent’s feints. A key to proper performance—in cognitive discrimination as in sports and performing arts—is to prevent things from disrupting this equilibrium by pulling the heart off balance, leading to blinkering and one-sided, erroneous execution.
要恰当地辨别，心必须保持一种平衡的、专注的状态，就像等待指挥家提示的独奏者，或应对对手假动作的篮球运动员一样。在认知辨别中，就像在体育和表演艺术中一样，恰当表现的关键是防止事物扰乱这种平衡，使心失去平衡，导致蒙蔽和片面的、错误的执行。

Thus the human heart can be compared to a pan of water. Place it upright and do not move it, and the sediment settles to the bottom and the clear water rises to the top. Then it is sufficient to see your beard and eyebrows and discern the fine patterns on your face. A breeze passing over it, the sediment moves below and the clear water is disturbed on top, and you cannot get even the general outline right. The heart too is like this. So guide it with pattern, cultivate it with clarity, and let nothing bias it. Then it will be sufficient to fix shì-fēi and settle doubts. If minor things pull it about, then externally one’s uprightness will be altered and internally the heart will be biased, and it will be insufficient to decide even gross patterns. (Xúnzǐ 21/54–58)
故人心譬如槃水，正错而勿动，则湛浊在下，而清明在上，则足以见须眉而察理矣。微风过之，湛浊动乎下，清明乱于上，则不可以得大形之正也。心亦如是矣。故导之以理，养之以清，物莫之倾，则足以定是非决嫌疑矣。小物引之，则其正外易，其心内倾，则不足以决粗理矣。（《荀子》21/54–58）

As this passage indicates, Xúnzǐ holds that proper “guiding” and “cultivating” in the arts of the heart can train us to avoid error. To discriminate shì-fēi and resolve confusing circumstances reliably, we can train ourselves to maintain an impartial, upright stance and an undisturbed, unbiased heart.
正如这段话所表明的，荀子认为，在心术方面进行恰当的 “引导” 和 “培养” 可以训练我们避免错误。为了可靠地辨别是非和解决困惑的情况，我们可以训练自己保持公正、正直的立场和不受干扰、无偏见的心。

The dào—in this case, Xúnzǐ’s preferred system of kinds (lèi 類), names (míng 名), ritual propriety (lǐ 禮), social roles (fēn 分), and associated duties (yì 義)—is the normative basis for such training. The dào provides an objective standard or “scale” (héng 衡，Xúnzǐ 21/29–30) to guide discrimination and action. An expert in a particular field, such as farming, commerce, or a craft, approaches that field on its own narrow terms. But an expert in the dào approaches all things inclusively, in terms of their relation to everything else (Xúnzǐ 21/51–52). Hence the gentlemen devotes himself wholly to mastering the dào and using it as a basis for examining things. Since he is focused fully on the dào, rather than particular interests or preferences, he is upright and unbiased; since he uses the dào as a basis for examining things, he is discerning in how he discriminates them into kinds. “Using impartial intentions to proceed with discerning sorting of kinds, one can manage the myriad things” (Xúnzǐ 21/52–53).
道 —— 在这种情况下，是荀子所偏好的类（种类）、名（名称）、礼（礼仪）、分（社会角色）和义（相关义务）的系统 —— 是这种训练的规范性基础。道提供了一个客观标准或 “衡”（尺度）（《荀子》21/29–30）来指导辨别和行动。某个特定领域的专家，如农业、商业或手工艺领域，会以其自身狭隘的术语来处理该领域。但道的专家会以包容性的方式处理所有事物，考虑它们与其他所有事物的关系（《荀子》21/51–52）。因此，君子全身心致力于掌握道，并以道为基础来考察事物。由于他完全专注于道，而非特定的兴趣或偏好，他是正直且无偏见的；由于他以道为基础来考察事物，他在将事物区分为不同种类方面是有洞察力的。“以仁心说，以学心听，以公心辨。不动乎众人之非誉，不治观者之耳目，不赂贵者之权势，不利传辟者之辞。故能处道而不贰，吐而不夺，利而不流，贵公正而贱鄙争，是士君子之辨也。”（《荀子》21/52–53）

Bias and error can thus be avoided by refraining from applying personal opinions or preferences and instead objectively weighing things according to the dào (Xúnzǐ 21/28–30). To do this reliably, however, we must recognize the dào and train ourselves in it by being xū 虛 (open, empty, or vacant), yī 壹 (focused or unified), and jìng 靜 (calm or still) (Xúnzǐ 21/34–35). By “openness,” Xúnzǐ means “not letting what one has already learned interfere with what is to be received” (Xúnzǐ 21/36–37): though the heart is constantly storing things in memory, it can remain open to new learning. By “focused,” he means “not letting that one interfere with this one” (Xúnzǐ 21/38). As we have seen, the nature of cognition for Xúnzǐ lies in discriminating things from each other. In doing so, he says, we implicitly recognize both things jointly, in relation to each other, and thus the heart is directed in two ways. Still, he says, we can also understand each thing on its own, without interference from how we discriminate it from other things. By “calm,” he means “not letting dreaming and fancies disrupt knowing” (Xúnzǐ 21/39). The heart is always in motion, Xúnzǐ thinks: during sleep, it dreams; while unoccupied, it wanders in idle fancy; while engaged, it plans. Yet we can also maintain a form of calm, through which we can avoid distraction by the imagination or irrelevant thoughts. By being open, says Xúnzǐ, one who seeks the dào may enter it; by being focused, one who undertakes to follow it may master it; by being calm, one who contemplates the dào can become discerning (Xúnzǐ 21/39–40). These psychological techniques, along with training in dào, supposedly provide a means of preventing blinkering or one-sidedness and thus error.
因此，通过不应用个人意见或偏好，而是根据道客观地权衡事物，可以避免偏见和错误（《荀子》21/28–30）。然而，要可靠地做到这一点，我们必须认识道，并通过 “虚”（开放、空虚或空无）、“壹”（专注或统一）和 “静”（平静或静止）来训练自己（《荀子》21/34–35）。荀子所说的 “虚”，意思是 “不以所已藏害所将受”（《荀子》21/36–37）：尽管心不断地将事物储存在记忆中，但它可以保持对新学习的开放。他所说的 “壹”，意思是 “不以夫一害此一”（《荀子》21/38）。正如我们所看到的，对荀子而言，认知的本质在于将事物彼此区分开来。他说，在这样做时，我们含蓄地共同认识到这两种事物及其相互关系，因此心有两个方向。不过，他说，我们也可以独立地理解每个事物，不受我们如何将它与其他事物区分开来的干扰。他所说的 “静”，意思是 “不以梦剧乱知”（《荀子》21/39）。荀子认为，心总是在运动：睡觉时，它会做梦；空闲时，它会胡思乱想；忙碌时，它会计划。然而，我们也可以保持一种平静，通过这种平静，我们可以避免被想象或无关的想法分散注意力。荀子说，虚，则入道；壹，则尽道；静，则悟道（《荀子》21/39–40）。这些心理技巧，连同对道的训练，据说提供了一种防止蒙蔽或片面性从而防止错误的方法。

6.2 Xúnzǐ on Illusion

6.2 荀子论幻觉

“Resolving Obscuration” explicitly discusses perceptual error, including error due to illusion. The conception of “obscuration” or “blinkering” (bì) shifts slightly in this discussion. Here, instead of attributing obscuration or blinkering primarily to psychological factors that bias the agent’s attention and discrimination, Xúnzǐ attributes it to external circumstances that interfere with normal perception, as when darkness hampers vision or distance obscures the size of things. His main claim is not that sense perception is misleading or unreliable, but that only a “fool” (Xúnzǐ 21/73) injudiciously discriminates whether things are “so” or not in potentially confusing circumstances. A cognitively competent agent recognizes and compensates for such circumstances and so avoids error. Part of our basic cognitive competence, for Xúnzǐ, is the ability to respond to “doubtful” conditions by withholding judgment accordingly: “Whenever in observing things there is doubt, if the heart within is not stable, then external things are unclear. Our thinking (lǜ 慮) being unclear,33 it is not yet permissible to fix ‘so’ or ‘not-so’” (Xúnzǐ 21/67–68). With appropriate care, error can be prevented even in confusing perceptual conditions. For ultimately it is due not to the conditions that temporarily obscure observation, but to how we employ the heart in response to them.
《解蔽》明确讨论了知觉错误，包括由幻觉引起的错误。在这一讨论中，“蔽”（遮蔽或蒙蔽）的概念略有变化。在这里，荀子没有将遮蔽或蒙蔽主要归因于使主体的注意力和辨别产生偏见的心理因素，而是将其归因于干扰正常知觉的外部环境，例如黑暗妨碍视觉或距离遮蔽事物的大小。他的主要主张不是感官知觉具有误导性或不可靠，而是只有 “愚者”（《荀子》21/73）才会在潜在的困惑环境中轻率地辨别事物是否 “如此”。有认知能力的主体会识别并补偿这种环境，从而避免错误。对荀子而言，我们基本认知能力的一部分是能够通过以下方式应对 “可疑” 条件相应地暂缓判断：“凡观物有疑，中心不定，则外物不清。吾虑不清，33 未可定然否也”（《荀子》21/67–68）。只要适当谨慎，即使在令人困惑的知觉条件下，也可以防止错误。因为错误最终不是由于暂时遮蔽观察的条件，而是由于我们如何运用心来应对这些条件。

33 Lǜ is “forethought,” or thought directed toward determining what judgment or action to undertake.
33 “虑” 是 “预先思考”，或旨在确定采取何种判断或行动的思考。

Xúnzǐ mentions seven types of circumstances in which in observation there are confusion and grounds for doubt (Xúnzǐ 21/68–73). First, there are familiar cases of perceptual illusion and hallucination. Someone walking in the dark may take a horizontal stone to be a crouching tiger or an upright tree to be a person following him, because “the darkness obscures his eyesight.” Pressing on our eyes, we take one thing to be two; covering our ears, we take a “mò mò” sound to be “xiōng xiōng,”34 because “the circumstances disrupt the sense organs.” From a mountaintop, oxen below look similar to sheep, because “the distance obscures their size.” From far below, tall trees on a mountain look similar to chopsticks, because “the height obscures their length.” The point of these examples is not that perception is deceptive or that appearances may fail to represent reality accurately. It is that we need to employ the heart carefully. An agent whose heart is stable and thinking is clear attends to the overall circumstances, not just to partial, yet misleading similarities between aspects of the situation and familiar things. “So from a mountaintop looking down at oxen, they are similar to sheep, but someone seeking sheep does not go down to lead them away… From the foot of a mountain looking up at trees, ten-meter trees are similar to chopsticks, but someone seeking chopsticks does not go up to break them off” (Xúnzǐ 21/71–72). Presumably, the norms and skills involved in discriminating things such as tigers, people, number, sounds, sheep, and chopsticks include provisions for discounting or compensating for the effects of unfavorable observational conditions. So an agent competent in these norms and skills performs properly even in such conditions.
荀子提到了七种在观察中存在困惑和怀疑理由的情况（《荀子》21/68–73）。首先，是常见的知觉幻觉和幻听案例。在黑暗中行走的人可能会把一块横放的石头当成蹲伏的老虎，或者把一棵直立的树当成跟在他后面的人，因为 “冥冥蔽其明也”。按压我们的眼睛，我们会把一个东西看成两个；捂住耳朵，我们会把 “漠漠” 的声音当成 “汹汹” 的声音，34 因为 “势乱其官也”。从山顶上看下面的牛，牛看起来像羊，因为 “远蔽其大也”。从下面很远的地方看山上的高树，高树看起来像筷子，因为 “高蔽其长也”。这些例子的要点不是知觉具有欺骗性，或者表象可能无法准确地代表实在。而是我们需要谨慎地运用心。心稳定且思维清晰的主体会关注整体环境，而不仅仅是情境的某些方面与熟悉事物之间的部分但具有误导性的相似性。“故从山上望牛者若羊，而求羊者不下牵也…… 从山下望木者，十仞之木若箸，而求箸者不上折也”（《荀子》21/71–72）。想必，在辨别老虎、人、数量、声音、羊和筷子等事物时所涉及的规范和技能，包括对不利观察条件的影响进行折扣或补偿的规定。因此，精通这些规范和技能的主体即使在这种条件下也能正确表现。

34 Some commentators suggest that the phrase mò mò 漠漠 refers not to a sound but to silence.
34 一些注释者认为，“漠漠” 一词不是指声音，而是指寂静。

The remainder of the examples confirms that Xúnzǐ’s theme is not that perception or observation is deceptive, but that potentially confusing circumstances call for special care and discretion. The further examples are intoxication, unreliable means, and incompetent testimony. A drunk will foolishly attempt to leap a wide ditch, taking it to be a mere gutter, or will stoop while exiting the city gate, taking it to be a low doorway, because “the alcohol disrupts his spirit” (Xúnzǐ 21/70). Here the issue is not perception, but how circumstances that disrupt normal functioning, as intoxication does, can cause massive failures of judgment. We do not discriminate whether we look attractive by our reflection in moving water, because “the water’s position is disturbed” (Xúnzǐ 21/72). Nor do we discriminate whether there are stars in the sky by asking the blind, because their “functional proficiency is confused” (Xúnzǐ 21/73).35 As with drunkenness, these would amount to attempting to discriminate things, as Xúnzǐ puts it, “by using confusion to resolve confusion.” In such cases, error is almost inevitable (Xúnzǐ 21/74).
其余的例子证实，荀子的主题不是知觉或观察具有欺骗性，而是潜在的困惑环境需要特别的谨慎和判断力。更多的例子是醉酒、不可靠的手段和无能的证词。醉汉会愚蠢地试图跳过一条宽阔的沟渠，把它当成一条普通的排水沟，或者在走出城门时弯腰，把城门当成一个低矮的门口，因为 “酒乱其神也”（《荀子》21/70）。这里的问题不是知觉，而是像醉酒这样扰乱正常功能的环境如何会导致严重的判断失误。我们不会通过流动水中的倒影来辨别自己是否有吸引力，因为 “水势玄也”（《荀子》21/72）。我们也不会通过询问盲人来辨别天空中是否有星星，因为他们的 “用精惑也”（《荀子》21/73）。35 与醉酒一样，这些都相当于试图正如荀子所说，“以惑解惑” 地辨别事物。在这种情况下，错误几乎是不可避免的（《荀子》21/74）。

35 Even in the case of the blind being unable to see the stars, Xúnzǐ’s explanation is that their “functional proficiency” (用精) is “confused,” not that they are in some way cut off from the world.
35 即使在盲人看不见星星的情况下，荀子的解释也是他们的 “用精”（功能熟练程度）被 “惑”（混淆）了，而不是他们在某种程度上与世界隔绝了。

For Xúnzǐ, then, error due to illusion is a subset of a more general sort of error resulting from an inept attempt to discriminate so from not-so in confusing circumstances when the heart is unstable and thinking unclear. The source of error lies in the agent’s incompetence or poor judgment, not the misleading nature of perception or appearance. When illusion occurs, the agent discriminates on the basis of some genuine similarity between his circumstances and models or exemplars of the kind in question, and thus is indisputably in touch with the world. In some respects, the outline of a horizontal stone may indeed resemble that of a tiger, or the shape of distant oxen in the valley below may indeed resemble that of nearby sheep. The problem is that these limited similarities are an insufficient basis for discriminating the stone as a tiger or the oxen as sheep. For the similarity is only partial, and there are other, dissimilar features to which a competent agent should also attend, even if “confusing” or “obscuring” circumstances make doing so more difficult than usual.36 If agents do attend to the overall circumstances, employing a “stable” heart and “clear” thinking, they will recognize that the similarity is insufficient and will avoid error. Illusion thus provides grounds for concluding that perception is reliable, not unreliable, provided the heart is employed properly.
因此，对荀子而言，由幻觉引起的错误是一种更一般错误的子集，这种错误源于在心不稳定且思维不清晰的情况下，在困惑的环境中试图笨拙地辨别是与非。错误的根源在于主体的无能或判断失误，而不是知觉或表象的误导性。当幻觉发生时，主体基于其环境与所讨论种类的模型或范例之间的某些真实相似性进行辨别，因此无可争议地与世界保持联系。在某些方面，横放的石头的轮廓可能确实像老虎的轮廓，或者山谷下方远处的牛的形状可能确实像附近的羊的形状。问题是这些有限的相似性不足以作为将石头辨别为老虎或牛辨别为羊的依据。因为这种相似性只是部分的，而且有能力的主体还应该关注其他不相似的特征，即使 “困惑” 或 “遮蔽” 的环境使这样做比平时更困难。36 如果主体确实关注整体环境，运用 “稳定” 的心和 “清晰” 的思维，他们会认识到这种相似性是不充分的，并会避免错误。因此，只要心被恰当运用，幻觉就为得出知觉是可靠的而非不可靠的这一结论提供了依据。

36 Xúnzǐ’s discussion of perceptually confusing circumstances does not explicitly apply part-whole rhetoric to explain error. The emphasis is instead on competently noting when one’s thinking is unclear or when some aspect of the circumstances produces “obscuration.” But a part-whole explanation can easily be extended to fit the examples. If we attend only to part of the situation—the rough similarity in shape between a stone and a tiger, as viewed in the dark, or that in size between oxen and sheep, as viewed from a distance—then we might mistakenly take a stone to be a tiger or oxen to be sheep. But if we attend to all aspects of the situation, including the darkness and the distance, error is unlikely.
36 荀子对知觉上令人困惑的环境的讨论没有明确应用部分 - 整体的修辞来解释错误。相反，重点在于胜任地注意到自己的思维何时不清晰，或环境的某些方面何时产生 “遮蔽”。但部分 - 整体解释可以很容易地扩展以适应这些例子。如果我们只关注情况的一部分 —— 在黑暗中看，石头和老虎在形状上的大致相似性，或者从远处看，牛和羊在大小上的相似性 —— 那么我们可能会错误地把石头当成老虎，把牛当成羊。但如果我们关注情况的所有方面，包括黑暗和距离，错误就不太可能发生。

7 Skepticism and Dream Arguments

7 怀疑论与梦的论证

Skeptical doctrines typically exist in symbiosis with a particular conception of knowledge, the skeptic questioning whether knowledge under that conception is attainable. As we would expect, then, the conception of knowledge as a form of ability or competence shapes the expression of skepticism in pre-Hàn thought. The very notion of an ability entails that its exercise is generally reliable, for otherwise we would not think of it as an ability. Moreover, what we take to be abilities must on the whole be fairly successful ways of interacting with the world, or we as a species would have perished long ago. If, with early Chinese epistemologists, we think of knowledge as a reliable ability to discriminate and respond to things, clearly we regularly succeed in discriminating and responding to things according to some fairly useful system of norms, for otherwise we would be unable to go about our daily lives. Given these points, in early Chinese thought there is little motivation for skepticism about whether sense perception presents the world accurately, whether we might know only appearances and not reality, whether our subjective ideas and beliefs correspond to the external world, or whether language accurately captures the nature of reality.37 All of these skeptical concerns ultimately arise from a representational theoretical framework that early Chinese thinkers do not share. Instead, in this intellectual context, skepticism takes the form of questioning whether our system of norms for discriminating and responding to things—in effect, our dào—is correct or appropriate, whether the claim that it is has any purchase, and whether language, fǎ (models), and other explicit guidelines genuinely have the capacity to guide us in following the dào reliably. These are the sorts of skeptical issues raised in the Dàodéjīng 道德經 and Zhuāngzǐ. That the dào we follow is genuinely a way of engaging with the world is not a matter of doubt. The question is rather whether there might be a more appropriate way—a more effective, unbiased, or comprehensive dào.
怀疑论学说通常与特定的知识概念共生，怀疑论者质疑在该概念下知识是否可获得。因此，正如我们所预期的，将知识视为一种能力或胜任力的概念，塑造了汉代以前思想中怀疑论的表现形式。能力这一概念本身就意味着其运用通常是可靠的，否则我们不会将其视为一种能力。此外，我们所认为的能力总体上必须是与世界互动的相当成功的方式，否则我们作为一个物种早就灭绝了。如果我们与中国早期认识论者一样，将知识视为可靠地辨别和回应事物的能力，那么显然我们会根据一些相当有用的规范体系，定期成功地辨别和回应事物，否则我们将无法进行日常生活。鉴于这些观点，在中国早期思想中，几乎没有动机去怀疑感官知觉是否准确地呈现世界，我们是否可能只知道表象而不知道实在，我们的主观观念和信念是否与外部世界相对应，或者语言是否准确地捕捉了实在的本质。37 所有这些怀疑论担忧最终都源于中国早期思想家所不认同的表征性理论框架。相反，在这种思想背景下，怀疑论采取的形式是质疑我们用于辨别和回应事物的规范体系 —— 实际上是我们的道 —— 是否正确或适当，关于它是正确的或适当的这一主张是否有任何依据，以及语言、法（模型）和其他明确的指导方针是否真正有能力指导我们可靠地遵循道。这些是《道德经》和《庄子》中提出的那种怀疑论问题。我们所遵循的道确实是一种与世界打交道的方式，这一点毋庸置疑。问题在于是否可能存在一种更适当的方式 —— 一种更有效、无偏见或更全面的道。

37 This statement should be qualified in one respect. Some passages in the Zhuāngzǐ question our ability to know the underlying causes or source of the things we observe—what is behind natural phenomena or the various attitudes we find within ourselves. For instance, the “Discourse on Evening Things Out” questions whether we can identify whatever it is that plays the “pipes of nature” and thus causes the activity of the myriad things (Zhuāngzǐ 2/9), whether we can know of a “ruler” or “master” within us that governs the other organs and from which our attitudes arise (Zhuāngzǐ 2/15–18), and whether we can know what ultimately causes our behaviour (Zhuāngzǐ 2/92–94). The gist of this line of skeptical questioning, however, is not that we know only appearance and not reality, but that our epistemic powers are fundamentally limited. In regard to certain questions, our ability to discriminate things cannot get off the ground, as there are no identifiable features to work with. On such issues, see too the exchange at 38 Zhuāngzǐ 17/20–24 and the dialogue at Zhuāngzǐ 25/59–82.
37 这一陈述在一个方面需要限定。《庄子》中的一些段落质疑我们了解所观察事物的潜在原因或来源的能力 —— 自然现象背后是什么，或者我们在自己内心发现的各种态度背后是什么。例如，《齐物论》质疑我们是否能识别出演奏 “天籁” 从而引起万物活动的东西（《庄子》2/9），我们是否能知道我们体内有一个 “统治者” 或 “主人” 支配着其他器官并从中产生我们的态度（《庄子》2/15–18），以及我们是否能知道我们行为的最终原因是什么（《庄子》2/92–94）。然而，这种怀疑论质疑的要点不是我们只知道表象而不知道实在，而是我们的认知能力从根本上是有限的。对于某些问题，我们辨别事物的能力无法开展，因为没有可识别的特征可供利用。关于这些问题，也参见《庄子》17/20–24 的对话和《庄子》25/59–82 的对话。

The claim that sense skepticism is not a concern in early Chinese thought might seem open to obvious counterexamples, for a pair of dream arguments figure prominently in the most important pre-Hàn skeptical text, Book 2 of the Zhuāngzǐ, the “Discourse on Evening Things Out 齊物論.” Under the influence of Descartes, contemporary readers may be inclined to assume that dream arguments inevitably concern the reliability of sense perception. In fact, however, the dream arguments in “Evening Things Out” concern change or transformation, not perception.38 Their main point is that how we apply action-guiding distinctions—examples include delight versus dislike, social ranks such as shepherd or noble, and what I am versus what I am not— depends partly on contingent, shifting circumstances. As circumstances change, we may come to draw these distinctions differently, for we may gain more information, change how we weigh different factors, or change our values. Thus we have two sorts of grounds for questioning whether our way of drawing such distinctions is conclusive: epistemic limitations due to our circumstances and potential shifts in our attitudes as circumstances change.
感官怀疑论在中国早期思想中不是一个关注点，这一主张似乎存在明显的反例，因为在汉代以前最重要的怀疑论文本《庄子》第二篇《齐物论》中，有一对梦的论证非常突出。在笛卡尔的影响下，当代读者可能倾向于认为梦的论证不可避免地涉及感官知觉的可靠性。然而，事实上，《齐物论》中的梦的论证涉及的是变化或转变，而非知觉。38 它们的主要观点是，我们如何应用指导行动的区分 —— 例如喜欢与不喜欢、牧羊人与贵族等社会等级，以及我是什么与我不是什么 —— 部分取决于偶然的、不断变化的环境。随着环境的变化，我们可能会以不同的方式做出这些区分，因为我们可能会获得更多信息，改变我们对不同因素的权衡，或者改变我们的价值观。因此，我们有两种理由质疑我们做出这种区分的方式是否具有决定性：由于我们的环境而产生的认知限制，以及随着环境变化我们态度的潜在转变。

38 On this point, see Fraser 2009b. Hansen has long argued that Zhuangist skepticism does not concern the reliability of sense perception (Hansen 1983: 92 and 183, fn. 80).
38 关于这一点，见弗雷泽 2009b。汉森长期以来一直认为，庄子的怀疑论与感官知觉的可靠性无关（汉森 1983：92 和 183，注 80）。

The first of the dream arguments questions how we apply evaluative, actionguiding distinctions—specifically, the distinction between delight and dislike. The text challenges even the most obvious application of the distinction: we delight in life and dislike death and so seek to preserve life and avoid death (Zhuāngzǐ 2/78–79). Perhaps the radical change in our circumstances that death brings would lead us to delight in death more than life, just as Lady Lì, who wept when married off to the King of Jìn, later came to delight in her pleasant life with him (Zhuāngzǐ 2/79–80). The experience of waking from a dream is cited as a vivid example of how evaluative attitudes can change with circumstances. Feeling delighted during a pleasant dream is justified. But if we awake to find ourselves in less pleasant circumstances, our attitudes change accordingly. The point is not that perception may be illusory or our access to the world unreliable. It is that how we draw value distinctions is as contingent as the judgments we make in a dream. We may feel confident about our judgments and then have them shift radically with an abrupt change in our situation (Zhuāngzǐ 2/81–83).
第一个梦的论证质疑我们如何应用评价性的、指导行动的区分 —— 具体来说，是喜欢和不喜欢之间的区分。文本甚至挑战了这种区分最明显的应用：我们喜欢生而不喜欢死，因此寻求保命避死（《庄子》2/78–79）。死亡带来的我们环境的彻底改变，或许会使我们喜欢死胜过喜欢生，就像丽姬嫁给晋君时哭泣，后来却喜欢上了和他在一起的愉快生活一样（《庄子》2/79–80）。从梦中醒来的经历被引用为评价态度如何随环境变化的生动例子。在梦中感到愉悦是合理的。但如果我们醒来发现自己处于不太愉快的环境中，我们的态度也会相应改变。要点不是知觉可能是虚幻的，或者我们对世界的接近不可靠。而是我们做出价值区分的方式，与我们在梦中做出的判断一样是偶然的。我们可能对自己的判断感到自信，然后随着我们处境的突然变化，这些判断会发生根本性的转变（《庄子》2/81–83）。

The second dream argument is the well-known story of Zhuāng Zhōu’s butterfly dream (Zhuāngzǐ 2/94–96). In the dream, Zhōu finds it obvious that he is a butterfly. On waking, he finds it equally obvious that he is Zhōu. Noticing how a change in circumstance—awakening—affected the obviousness of whether he was a butterfly or Zhōu, he wonders whether he has now finally discriminated his identity correctly, for circumstances could change yet again. Between Zhōu and the butterfly there is surely a distinction, the text affirms. But Zhōu lacks grounds to determine conclusively which side of the distinction he is on. This, the text says, illustrates how things change (wù huà 物化，2/96). The point of the story is thus to question whether discriminations we make in one set of circumstances hold for all, given that things inevitably change. It is not to doubt whether perception reliably represents reality.
第二个梦的论证是著名的庄周梦蝶的故事（《庄子》2/94–96）。在梦中，周明显觉得自己是一只蝴蝶。醒来后，他同样明显地觉得自己是周。注意到环境的变化 —— 醒来 —— 如何影响他是蝴蝶还是周的明显性，他怀疑自己现在是否终于正确地辨别了自己的身份，因为环境可能会再次变化。文本确认，周和蝴蝶之间肯定有区别。但周没有依据来最终确定自己属于这个区别的哪一边。文本说，这说明了事物是如何变化的（物化，2/96）。因此，这个故事的要点是质疑，鉴于事物不可避免地会变化，我们在一组环境中做出的辨别是否适用于所有环境。它不是为了怀疑知觉是否可靠地表征实在。

8 Conclusion

8 结论

The skeptical force of the argument from illusion grows out of a conception of mind and knowledge that allows for the possibility of a comprehensive or persistent gap between reality and our perception of it. A core feature of such a conception, I have suggested, is a broadly representational understanding of mind and knowledge. By contrast, early Chinese theories of mind and knowledge explain cognitive functions in terms of abilities or competence, not representation. Such theories leave no conceptual room for the notion of a gap between the agent and the world. An agent who commits a cognitive error, including perceptual error, is in contact with the world, just as one who burns the soufflé or hooks a golf drive into the trees is. The problem lies in faulty performance, not the inherently deceptive nature of perception or the unknowable character of reality. The mistaken agent has simply failed to pay attention to the right things or respond to them in the right way. Even in the case of perceptual error due to illusion, the agent is not regarded as misled by appearances or sense impressions that misrepresent reality. Rather, the agent is seen as in touch with the world and usually right about something, namely that whatever feature prompted the erroneous discrimination is indeed similar, at least partially, to a paradigm of the kind of thing the agent took it to be.

来自幻觉的论证的怀疑论力量源于一种心灵和知识的概念，这种概念承认实在与我们对它的知觉之间可能存在全面或持久的鸿沟。我认为，这种概念的一个核心特征是对心灵和知识的广泛表征性理解。相比之下，中国早期的心灵和知识理论根据能力或胜任力而非表征来解释认知功能。这种理论没有为主体与世界之间的鸿沟概念留下概念空间。犯认知错误（包括知觉错误）的主体与世界保持联系，就像烤糊舒芙蕾或把高尔夫球开到树丛中的人一样。问题在于不当的表现，而非知觉固有的欺骗性或实在的不可知特性。犯错的主体只是未能关注正确的事物或以正确的方式对它们做出回应。即使在由幻觉引起的知觉错误的情况下，主体也不被视为被歪曲实在的表象或感官印象所误导。相反，主体被视为与世界保持联系，并且通常在某些方面是正确的，即促使错误辨别的任何特征确实至少部分地类似于主体所认为的那种事物的范例。

As the preceding sections have shown, early Chinese texts present three sorts of explanations of cognitive error, including perceptual error. The first is an explanation of just what error is: it is confusion or failure in discriminating and responding to things. The second is a series of explanations of various psychological and environmental factors that can cause error. For a competent agent, who normally knows what he perceives, perceptual error is due to occasional performance failure arising from psychological factors such as carelessness, distraction, obsession, or nervousness, sometimes in conjunction with environmental factors such as indistinct or confusing observational conditions. The third explanation is a more general, partly metaphorical interpretation of what happens when such factors interfere with correct discrimination: the agent responds to some features of the circumstances while overlooking other relevant features, in effect responding to an unrepresentative or insufficient part of the situation rather than competently coping with the whole. At this general, abstract level, instead of a distinction between appearance and reality, early Chinese theorists explain error by appeal to a distinction between part and whole. Although only Xúnzǐ elaborates such a general, part-whole explanation in much detail, I have tried to show that a part-whole view coheres well with Mohist epistemology and with other early discussions of error, such as those in The Annals of LÜ Buwei.

正如前面各节所表明的，中国早期文献对包括知觉错误在内的认知错误提出了三种解释。第一种是解释错误是什么：它是辨别和回应事物时的困惑或失败。第二种是对可能导致错误的各种心理和环境因素的一系列解释。对于通常知道自己所感知的事物的有能力的主体而言，知觉错误是由于偶尔的表现失误，这种失误源于心理因素，如粗心、分心、执念或紧张，有时还与环境因素（如模糊不清或令人困惑的观察条件）相结合。第三种解释是对当这些因素干扰正确辨别时所发生情况的一种更一般的、部分隐喻性的解释：主体对环境的某些特征做出回应，而忽略了其他相关特征，实际上是对情境中不具代表性或不充分的部分做出回应，而非胜任地应对整体。在这个一般的、抽象的层面上，中国早期理论家不是通过表象与实在的区分来解释错误，而是通过部分与整体的区分。尽管只有荀子详细阐述了这种一般的部分 - 整体解释，但我试图表明，部分 - 整体观点与墨家认识论以及其他早期对错误的讨论（如《吕氏春秋》中的讨论）高度一致。

For pre-Hàn thinkers, the situation as a whole always provides sufficient features for the competent, conscientious agent to avoid perceptual error. We are expected to notice whether our circumstances are “confusing” and compensate accordingly, either by employing care and discretion in discrimination or by temporarily withholding judgment. In the rare case when we cannot discriminate correctly, even after investigation, we can at least determine the nature of the “obscuration” that prevents us from doing so. If we perform competently, then, we will either discriminate correctly or at minimum avoid error. Moreover, in pre-Hàn thought, agents are conceived of as performers of a dào, a continuing path or course of conduct. Instances of perceptual discrimination are thus seen as parts of an ongoing course of activity, not discrete, isolated judgments that stand or fall on their own. Hence if we do err, we can expect to discover and correct the mistake in due course. The foolish shepherd who discriminates distant oxen as sheep and hikes down to lead them back to the flock will revise his discrimination soon enough. The object of skeptical doubt in pre-Hàn thought is not whether we ever succeed in interacting with the world competently—and thus obtain knowledge of it—but whether our norms for interacting with it are uniquely or absolutely correct. Perhaps other norms might be equally or more justified, for other people or creatures, if not us. That we can achieve competence is not in doubt. The question is whether our form of competence is the only or the most appropriate one.
对汉代以前的思想家而言，情境作为一个整体总是为有能力、认真的主体提供足够的特征来避免知觉错误。我们应该注意到我们的环境是否 “令人困惑”，并相应地进行补偿，要么在辨别时运用谨慎和判断力，要么暂时暂缓判断。在极少数情况下，即使经过调查我们也无法正确辨别，我们至少可以确定阻碍我们这样做的 “遮蔽” 的性质。因此，如果我们表现胜任，我们要么会正确辨别，要么至少会避免错误。此外，在汉代以前的思想中，主体被视为 “道” 的践行者，“道” 是一条持续的路径或行为过程。因此，知觉辨别实例被视为一个持续活动过程的一部分，而非孤立的、独立的、自行成立或不成立的判断。因此，如果我们确实犯了错误，我们有望在适当的时候发现并纠正错误。那个把远处的牛误判为羊并下山去把它们赶回羊群的愚蠢牧羊人，很快就会修正他的判断。汉代以前思想中怀疑论怀疑的对象，不是我们是否曾经成功地与世界进行过胜任的互动 —— 从而获得对它的知识 —— 而是我们与世界互动的规范是否是唯一或绝对正确的。或许，即使对我们而言不是，对其他人或生物而言，其他规范也可能同样或更有正当性。我们能够获得胜任力这一点是毋庸置疑的。问题是我们的胜任力形式是否是唯一或最适当的。

Acknowledgements This article is a revised and expanded version of material first presented in Fraser 1999. Earlier versions were presented in February 2009 at a seminar in the Department of Philosophy, University of Hong Kong, and in May 2009 at “New Directions in Chinese Philosophy,” a conference celebrating the 60th anniversary of the Department of Philosophy of the Chinese University of Hong Kong. I thank participants in both events for many stimulating questions. For helpful criticisms and suggestions, I am grateful to an anonymous referee and especially to Jane Geaney, whose comments prompted numerous revisions.
致谢 本文是对首次发表于弗雷泽 1999 年的材料的修订和扩展版本。早期版本于 2009 年 2 月在香港大学哲学系的一次研讨会上，以及 2009 年 5 月在庆祝香港中文大学哲学系成立 60 周年的 “中国哲学新方向” 会议上进行了展示。我感谢这两次活动的参与者提出的许多富有启发性的问题。感谢一位匿名评审员，特别是简・吉尼提出的有益批评和建议，他们的评论促使我进行了多次修改。

References

参考文献

• Cua, A. S. 1985. Ethical Argumentation: A Study in Hsun Tzu’s Moral Epistemology. Honolulu: University of Hawaii Press.
  匡安乐，1985，《伦理论证：荀子道德认识论研究》，檀香山：夏威夷大学出版社。

• Dretske, Fred. 1981. Knowledge and the Flow of Information. Cambridge, MA: MIT Press.
  弗雷德·德雷斯克，1981，《知识与信息的流动》，马萨诸塞州剑桥：麻省理工学院出版社。

• Fraser, Chris. 1999. Similarity and Standards: Language, Cognition, and Action in Chinese and Western Thought. Ph.D. dissertation, University of Hong Kong.
  克里斯·弗雷泽，1999，《相似性与标准：中西方思想中的语言、认知与行动》，香港大学博士学位论文。

• _____. 2006. “Zhuangzi, Xunzi, and the Paradoxical Nature of Education.” Journal of Chinese Philosophy 33.4: 529–42.
  _____，2006，《庄子、荀子与教育的悖论性质》，《中国哲学杂志》33.4：529–42。

• _____. 2009a. “Mohist Canons.” In The Stanford Encyclopedia of Philosophy (Summer 2009 Edition), edited by Edward Zalta. http://plato.stanford.edu/archives/sum2009/entries/mohist-canons/.
  _____，2009a，《墨家辩经》，载于爱德华·扎尔塔编《斯坦福哲学百科全书》（2009 年夏季版），http://plato.stanford.edu/archives/sum2009/entries/mohist-canons/。

• _____. 2009b. “Skepticism and Value in the Zhuāngzǐ.” International Philosophical Quarterly 49.4: 439– 57.
  _____，2009b，《庄子中的怀疑论与价值》，《国际哲学季刊》49.4：439–57。

• Geaney, Jane. 2002. On the Epistemology of the Senses in Early Chinese Thought. Honolulu: University of Hawaii Press.
  简·吉尼，2002，《中国早期思想中的感官认识论》，檀香山：夏威夷大学出版社。

• Graham, A. C. 1989. Disputers of the Tao. La Salle: Open Court.
  A.C.格雷厄姆，1989，《道的论争者》，拉萨尔：开放法院出版社。

• _____. 2003. Later Mohist Logic, Ethics and Science. Reprint. Hong Kong: Chinese University Press.
  _____，2003，《后期墨家逻辑、伦理与科学》，重印本，香港：香港中文大学出版社。

• Hansen, Chad. 1983. Language and Logic in Ancient China. Ann Arbor: University of Michigan Press. . 1992. A Daoist Theory of Chinese Thought. New York: Oxford.
  查德·汉森，1983，《古代中国的语言与逻辑》，安阿伯：密歇根大学出版社。，1992，《道家思想的中国理论》，纽约：牛津大学出版社。

• Harbsmeier, Christoph. 1993. “Conceptions of Knowledge in Ancient China.” In Epistemological Issues in Classical Chinese Philosophy. Edited by Hans Lenk and Gregor Paul. Albany: SUNY Press.
  克里斯托夫·哈布斯迈尔，1993，《古代中国的知识概念》，载于汉斯·伦克和格雷戈尔·保罗编《中国古典哲学中的认识论问题》，奥尔巴尼：纽约州立大学出版社。

• _____. 1998. Science and Civilisation in China, Vol. 7, Part 1: Language and Logic. Cambridge: Cambridge University Press.
  _____，1998，《中国科学技术史》第 7 卷第 1 部分：《语言与逻辑》，剑桥：剑桥大学出版社。

• Knoblock, John. 1994. Xunzi: A Translation and Study of the Complete Works, vol. 3. Stanford: Stanford University Press.
  约翰·诺布洛克，1994，《荀子：全集翻译与研究》第 3 卷，斯坦福：斯坦福大学出版社。

• Knoblock, John, and Jeffrey Riegel, tr. 2000. The Annals of Lü Buwei. Stanford: Stanford University Press.
  约翰·诺布洛克、杰弗里·里格尔译，2000，《吕氏春秋》，斯坦福：斯坦福大学出版社。

• Lǐ, Díshēng 李滌生. 1979. Collected Explications of Xúnzǐ 荀子集釋. Taipei 台北: Xuéshēng 學生.
  李滌生，1979，《荀子集釋》，台北：學生書局。

• McDowell, John. 1996. Mind and World. Cambridge, Ma.: Harvard University Press.
  约翰·麦克道威尔，1996，《心灵与世界》，马萨诸塞州剑桥：哈佛大学出版社。

• _____. 1998. “Knowledge and the Internal.” In McDowell, Meaning, Knowledge, and Reality. Cambridge, MA: Harvard University Press.
  _____，1998，《知识与内在》，载于麦克道威尔《意义、知识与实在》，马萨诸塞州剑桥：哈佛大学出版社。

• Millikan, Ruth Garrett. 1987. Language, Thought, and Other Biological Categories. Cambridge, MA: MIT Press.
  露丝·加勒特·米利肯，1987，《语言、思想与其他生物学范畴》，马萨诸塞州剑桥：麻省理工学院出版社。

• Mòzǐ. 1986. A Concordance to Mòzǐ 墨子引得. Harvard-Yenching Institute Sinological Index Series, Supplement no. 21 (reprint). Shànghǎi 上海: Shànghǎi Gǔjí 上海古籍.
  《墨子》，1986，《墨子引得》，哈佛-燕京学社汉学索引系列补编第 21 号（重印本），上海：上海古籍出版社。

• Munro, Donald. 1969. The Concept of Man in Early China. Stanford: Stanford University Press.
  唐纳德·芒罗，1969，《早期中国的人的概念》，斯坦福：斯坦福大学出版社。

• Robins, Dan. 2008. “Xunzi.” In The Stanford Encyclopedia of Philosophy (Fall 2008 Edition). Edited by Edward Zalta. http://plato.stanford.edu/archives/fall2008/entries/xunzi/.
  丹·罗宾斯，2008，《荀子》，载于爱德华·扎尔塔编《斯坦福哲学百科全书》（2008 年秋季版），http://plato.stanford.edu/archives/fall2008/entries/xunzi/。

• Roetz, Heiner. 1993. “Validity in Zhou Thought.” In Epistemological Issues in Classical Chinese Philosophy. Edited by Hans Lenk and Gregor Paul. Albany: SUNY Press.
  海纳·罗茨，1993，《周代思想中的有效性》，载于汉斯·伦克和格雷戈尔·保罗编《中国古典哲学中的认识论问题》，奥尔巴尼：纽约州立大学出版社。

• Sivin, Nathan. 1995. “State, Cosmos, and Body in the Last Three Centuries B.C.” Harvard Journal of Asiatic Studies 55.1: 5–37.
  内森·席文，1995，《公元前三世纪的国家、宇宙与身体》，《哈佛亚洲研究杂志》55.1：5–37。

• Watson, Burton, tr. 1963. Hsun Tzu: Basic Writings. New York: Columbia University Press.
  伯顿·沃森译，1963，《荀子：基本著作》，纽约：哥伦比亚大学出版社。

• Xúnzǐ. 1966. A Concordance to Xúnzǐ 荀子引得. Harvard-Yenching Institute Sinological Index Series, Supplement no. 22. Cambridge, Ma.: Harvard University Press.
  《荀子》，1966，《荀子引得》，哈佛-燕京学社汉学索引系列补编第 22 号，马萨诸塞州剑桥：哈佛大学出版社。

• Zhuāngzǐ. 1956. A Concordance to Zhuāngzǐ 莊子引得. Harvard-Yenching Institute Sinological Index Series, Supplement no. 20. Cambridge, MA: Harvard University Press.
  《庄子》，1956，《庄子引得》，哈佛-燕京学社汉学索引系列补编第 20 号，马萨诸塞州剑桥：哈佛大学出版社。

---

via：

Dao (2011) 10:127–148
DOI 10.1007/s11712-011-9206-5