How Zen Became Chan – T. Griffith Foulk

Return to the main conference page.

Return to the conference abstracts page.

Work in progress, not for citation.
No part of this content can be used, distributed, and reproduced without written consent from the author.


Histories of Chan (Zen) 

by T. Griffith Foulk



The Scope and Aims of this Book

This book, Histories of Chan (Zen), focuses on the Chan School of Buddhism in China, from the first appearance of precursory elements in the Tang dynasty (618-906) to the flourishing of the school in the Song (960-1279) and Yuan (1280-1368) dynasties. Chan 禪, better known in English by the Japanese pronunciation Zen, has occasioned much scholarly research and publication from the late nineteenth century down to the present, but interest in it is not merely academic. Many Buddhist monks, nuns and lay followers today in various parts of East Asia (wherever Chinese cultural influence has been strong), and some in the West, as well, regard themselves and the organizations they belong to as heirs to the Chan School of medieval China.[1]

The mythology of the Chan Lineage (Chanzong 禪宗), embodied in a distinctive set of texts and rituals, was transmitted to Japan in conjunction with the latest in elite Chinese Buddhist monastery organization and operation, and a large dose of literati culture, during the Kamakura period (1185-1333). That provided the foundation for what became known as the Zen School (J. Zenshū 禪宗) of Buddhism.[2] The Sōtō and Rinzai denominations of Zen in Japan today, both of which trace their heritage directly back to Song China,[3] together comprise about 20,000 temples, 22,000 ordained members of the clergy, and 3,000,000 lay followers.[4] During the first century of the Edo period (1600-1868), Chan Lineage monks from the continent were invited to Japan and given the means to reproduce forms of Buddhist monastic practice that had evolved in Ming dynasty (1368-1644) China. The group of Chan (Zen) masters and network of monasteries that resulted was known as the Ōbaku school (J. Ōbakushū 黄檗宗), which today has about 450 temples and 350,000 lay followers.[5]

In present day South Korea, the Chogye order of celibate monks,[6] which regards itself as heir to the Korean Sǒn tradition,[7] dominates the modest Buddhist monastic establishment.[8] Although it is a denomination that was reconstructed in the aftermath of the Japanese rule of Korea (1905-1945), it draws much of its inspiration from the Chan School of the Song and Yuan, as interpreted and transmitted by the Korean monks Chinul 知訥 (1158-1210) and T’aego Pou 太古普愚 (1301-1382). Chinul’s connection to Chinese Chan was mainly through texts,[9] but T’aego visited China from 1346 to 1348, was recognized as a dharma heir in the Linji Chan Lineage, and returned to Korea with “rules of purity” (qinggui 清規, K. chŏnggyu) for the organization and operation of Buddhist monasteries that were closely associated with the Chan tradition.

In Vietnam, a Thiến School of Buddhism based on the Song Chan model was established with royal patronage early in the Trân dynasty (1225-1400).[10] In the late seventeenth century, a reform movement within the Vietnamese sangha established the Lâm-te School of Thiến,[11] which in its meditation practices, monkish etiquette, and religious rites was (like the Ōbaku School in Japan) patterned after the elite monastic Buddhism of Ming China. The Lâm-te tradition has survived to the present day in Vietnam and has given rise to other Thiến reform and revival movements that have sought legitimacy by emphasizing their roots in medieval Chinese Chan.

In China itself, Neo-Confucianism overshadowed Buddhism as the preferred religion of the educated elites (the so-called literati class) during the Ming dynasty, a time when the Buddhist monastic institution as a whole suffered a loss of prestige and patronage. During the seventeenth century, nevertheless, there was a revival of Chan (as that had been conceived and chronicled in texts dating from the Song and Yuan) that swept through the upper echelons of the Buddhist monastic community and was actively encouraged and embraced by a number of literati as well.[12] After the fall of the Ming and the establishment of the Qing dynasty (1644-1912), the revival movement continued to flourish with the support of the Manchu rulers and, as noted above, its influence was felt in Japan and Vietnam as well as China. During the Qing, the designation “Chan” came to refer loosely to the form of elite monastic Buddhism that was associated with the language, culture, and history of the Han Chinese, in contradistinction to the types of Buddhism (also sponsored by the Qing court) that derived from the Tibetan cultural milieu. Prior to the communist revolution of 1949, when there was a much larger community of Buddhist monks and nuns in mainland China than at present, virtually all of them belonged to Chan tonsure lineages (either the Linji or Caodong).[13]

With one major exception, this book does not deal directly with any of the developments that took place outside of China, or within China after the fourteenth century: it focuses mainly on the background and history of the Chan tradition in the Tang, Song, and Yuan dynasties. Like any work of academic historiography, however, the book builds on and takes issue with previously published studies of its subject. Thus — and this is the aforementioned exception — in the process of critiquing the large body of modern scholarship that treats the history of Chan Buddhism in medieval China, I do focus directly on the origins and development of the modern field of Zen studies (J. Zengaku 禪學).[14] I trace the roots of that field in the intellectual and institutional history of Chinese and Japanese Buddhism and explain how it came to take the shape that it has.

The modern field of Zen studies was founded in late nineteenth and early twentieth-century Japan by scholarly Zen monks and lay followers who, embracing principles and methods of scientific philology and historiography that had recently been introduced from the West, used those to revisit traditional histories of the Chan and Zen lineages that dated from the Song and later. Those pioneers compiled comprehensive collections and critical editions of primary source materials, produced basic reference works such as dictionaries and bibliographies, and wrote numerous monographs as well as some sweeping historical surveys. Without their prodigious output of academic publication, the field of Zen studies as we know it today would never have come into existence.

The field of Zen studies, of course, has undergone many changes since its inception. The discovery in Dunhuang (a city in western China) of manuscripts dating from the Tang Dynasty that disagree with traditional (Song Dynasty and later) accounts of the early Chan Lineage forced Japanese scholars, beginning in the late 1920s, to completely rethink the received history of that lineage. Around the same time, the field began to be joined by a few Chinese and Western scholars, and since the end of the Second World War it has become increasingly international. Considerable academic research has now been published in English, French, German, Chinese, and Korean as well as Japanese. Not all of the authors have been apologists for the Zen tradition, and a some have employed critical methodologies (e.g. deconstruction, Marxist analyses, etc.) that hold nothing of the tradition sacred.

Still, most of the academic research on the history of Chinese Chan that has taken place since the advent of Zen studies in the late nineteenth century has been stimulated by the interests and fostered by the support of contemporary religious communities: Buddhist organizations in Japan and elsewhere, including some that operate universities and institutes for academic research, that look back to the Chan School of medieval China as the source of their own traditions.

There is nothing wrong with that in principle. In my view, the notion of a purely objective or disinterested historiography is an oxymoron, for in theory there is no limit to the number of things that can be shown on the basis of sound evidence to have “really happened” in the past, and even the most scrupulously critical and objective (i.e. methodologically rigorous) of historians focus only on the particular things that matter to them. Bias of that sort is not only unavoidable in critical historiography, it is the life-blood of the entire endeavor.

Nevertheless, it is my contention that the modern field of Zen studies, taken as a whole, has suffered from two kinds of correctable bias that may be traced back to its origins in the Zen schools of late nineteenth and early twentieth century Japan. The first is a lack of critical distance from certain traditional religious beliefs that are the proper objects (qua doctrine or ideology) of historical research, but should not be employed as historiographical categories or accepted as names for empirically verifiable phenomena. The prime example is the concept of a spiritual “lineage” (zong 宗) formed by the unbroken transmission (chuan 傳) from masters to disciples of an avowedly signless (wuxiang 無相) dharma (fa 法).[15]

The second kind of bias is the tendency to valorize a “pure Chan” (J. junsui Zen 純粋禪) that is supposed to have existed in the “golden age” of the Tang Dynasty, and to depict any historically verifiable features of the later Chan and Zen schools that do not comport with that idealized vision as evidence of degeneration — a “syncretistic” embrace of extraneous elements — in the tradition. The latter bias, which involves the projection onto an imagined “original Chan” (J. shoki Zen 初期禪) of secular and humanistic values learned from the West, arose in response to political and ideological pressures that were endemic in Japan in the decades before and after the turn of twentieth century. At that time, Japanese Buddhism as a whole was under severe attack for its “superstitious” beliefs and ritual practices. Zen was held up as a kind of Buddhism that had originally (in the Tang) been free from such defects, and thus was ideally suited to meet the spiritual needs of people in the modern, scientific age.

As a result of these two kinds of bias and other methodological shortcomings, I further contend, most modern scholarship has failed to define its declared object of historical study — Chan (Zen) — in a clear and unambiguous manner.

As is well known to anyone with an interest in Chan, Sǒn, Thiến or Zen, traditional accounts of the history of Chan posit a spiritual lineage (zong 宗) of twenty-eight Indian and six Chinese ancestral teachers (zushi 祖師) that began when Śākyamuni Buddha (Shijiamouni fo 釋迦牟尼佛) transmitted his awakening — his “sublime mind of nirvāṇa” (niepan miaoxin 涅槃妙心) — directly to one of his disciples, the monk Mahākāśyapa (Mohejiashe 摩訶迦葉), in what is characterized as a “separate transmission apart from the teachings” (jiaowai biechuan 教外別傳) that did “not rely on words or letters (buli wenzi 不立文字). Mahākāśyapa, the first ancestor of the lineage in India, later transmitted the “mind-dharma” (xinfa 心法) of the Buddha to Ānanda (Anan 阿難), who became the second ancestor. The mind-dharma, also called the “buddha-mind” (foxin 佛心), was then handed down from master to disciple through the generations until it reached Bodhidharma (Putidamo 菩提達摩 or 菩提達磨), the twenty-eighth ancestor. Bodhidharma famously “came from the west” (xilai 西來), from India to China, becoming the first ancestor (chuzu 初祖) of the Chan Lineage in that country. The mind-dharma, it is said, was subsequently transmitted down through five more generations of Chinese ancestral teachers until it reached Huineng 慧能, the Sixth Ancestor (Liuzu 六祖), whose succession to the lineage is celebrated in the Platform Sūtra (Tanjing 壇經).[16] Traditional accounts depict a ramification of the Chan family tree in the generations following Huineng, resulting in five main branches, also known as the “five houses” (wujia 五家). Two of those branches, the Linji (J. Rinzai) 臨濟 and Caodong (J. Sōtō) 曹洞, are the lineages through which all followers of the Chan, Sǒn, Thiến and Zen schools today trace their spiritual heritage.

Even before they were confronted with hard textual evidence that undermined the historicity of this traditional account of the Chinese instantiation of the Chan Lineage, some of the Japanese scholars who pioneered the modern field of Zen studies had begun to suspect that the story of the lineage in India was basically a myth of Chinese origin. Because they had access to Buddhist literature written in Pāli and Sanskrit — sūtras, śāstras, and vinaya texts — that treated the life of Śākyamuni Buddha, the careers of his leading disciples, and the teachings of other Indian monks such as Nāgārjuna (Longshu 龍樹) who appeared in the Chan Lineage records, modern scholars gradually came to realize that the detailed biographies of the Indian ancestral teachers found in Chan records dating from the Song and Yuan dynasties could not be corroborated by any Indian sources, and thus had to be hagiographies that had been produced in China. Then, with the discovery of Dunhuang manuscripts that contained inconsistent, competing accounts of Bodhidharma’s Lineage in China (e.g., there are three different figures identifed as the “sixth ancestor” in those Tang sources), modern scholars had to face the fact that the traditional (Song and later) account of the early Chan Lineage in China was also highly unreliable, if not entirely fictional.

Those discoveries, coming as they did in the early decades of the twentieth century, led modern scholars to begin thinking about the history of Chan in terms other than the traditional category of “lineage” (zong 宗, J. shū). Once the “Lineage of Bodhidharma” was exposed (in at least some of its iterations) as a mythological entity, it would no longer do to rely entirely on traditional records of “dharma transmission” (chuanfa 傳法) to define the Chan movement that was ostensibly founded in China by Bodhidharma. Scholars thus felt the need to find some empirically verifiable teachings, practices or institutional arrangements that they could use to characterize the Chan School, to distinguish its various branches, and to trace its historical development. As a result, a number of theories concerning the essential nature of the early Chan movement were put forth, but no concensus on those matters emerged. Some scholars, for example, characterized Chan as the school that specialized in the practice of dhyāna or “meditation,” while others defined it as a radical, sectarian movement that rejected the practice of meditation (and most other aspects of Indian Buddhism) and championed the doctrine of “sudden awakening” (dunwu 頓悟).

However, despite the various attempts that modern scholars have made to define the early Chan School in terms of a concrete, empirically verifiable set of beliefs, practices, or institutional arrangements, few have seen fit to entirely abandon the traditional view of Chan as a fraternity of spiritually awakened men linked by the esoteric bond of “mind-to-mind” dharma transmission. All have continued to regard the membership of the Chan School — throughout its history in the Tang, Song and Yuan dynasty China — as comprising virtually the same individuals as the set of ancestral teachers who are named in traditional (Song and later) genealogies of the Chan Lineage. In short, modern scholars have never actually applied their new definitions or characterizations of the Chan School in a thoroughgoing manner, one that would override the old conceptual model of a lineage of dharma transmission and result in a membership that looks quite different. If, for example, Chan is defined as the “meditation school,” then the most famous and influential teacher of meditation practices in all of Chinese Buddhist history — Tiantai Zhiyi 天台智顗 (538-597) — should be considered a member. If, on the other hand, Chan is taken to be the “school of sudden awakening,” then it is hard to see how the Indian monk Bodhidharma, whose teachings are described in Daoxuan’s 道宣 (596-667) Additional Biographies of Eminent Monks (Xu gaoseng zhuan 續高僧傳), could possibly be a member, let alone the founder of the school in China. This kind of sloppy definition of the object of study is one of the fundamental shortcomings of the modern scholarship that this book is intended to redress.

What I present in this book is a radical rethinking and reorganization of the field of Zen studies. My aim is not simply to correct or revise particular aspects of the modern scholarship, although I do a fair amount of that. More importantly, I wish to provide a new conceptual framework within which all the existing pieces of a very complicated historical puzzle — a diverse set of stories about the origins, development, and essential characteristics of Chan Buddhism that have been told for different reasons at different times and places — can be sorted out and related to one another in a manner that makes sense and is consistent with all the evidence that we have today.

To that end, I divide my approach to the history of Chan into five distinct lines of inquiry, devoting a separate part to each:

Part One: A Lexical History of the Word “Chan
Part Two: Traditional Histories of the Chan Lineage
Part Three: Modern Histories of the Chan Lineage/School
Part Four: Proto-Histories of the Chan Lineage
Part Five: An Institutional History of the Chan School

The titles of these parts are not self-explanatory. That is to say, they do not provide sufficient information for the reader to deduce their respective contents or to grasp my reasons for presenting five different histories of Chan instead of just one. In the remaining sections of this Introduction, therefore, I define all of the terms used in these titles, elucidate the various objects and methods of study represented in each of the five parts, and explain why I have not written a straightforward “history of Chan Buddhism,” as if that were an unambiguous thing that could be treated in the singular.


Principles of Definition and Historical Inquiry


Matters of definition are a central concern of the present book and a theme that recurs throughout all five parts. The two most basic questions that inform my entire project are: (1) how has Chan been defined as an object of historical study in the past, and (2) how should Chan be defined as an object of historical study from now on?

Those questions had not occurred to me in 1976 when, after graduating from college and spending a few years in a Rinzai Zen training monastery (J. senmon dōjō 專門道場, a.k.a. sōdō 僧堂) in Japan, I entered graduate school at the University of Michigan and began the academic study of Buddhism. The questions only came up when I started to research the history of the Japanese Zen monastic institution, which had originally been imported from Song- and Yuan-dynasty China and was said to have evolved directly from a system of training first established in the Tang Dynasty by Chan master Baizhang Huaihai 百丈懐海 (749-814). As I read the traditional accounts of Baizhang’s rules that appear in Song histories of the Chan Lineage and went looking for corroborating evidence in texts dating from the Tang, it began to dawn on me just how ambiguous the word chan is across that range of Chinese Buddhist literature. It also became apparent that while modern scholars have adhered to the traditional (Song and later) genealogical lists of Buddhist monks who belonged to the Chan Lineage as a basic frame of reference, they have not agreed on what the word “Chan” means in the name of that lineage. A number of different doctrines, practices, and institutional arrangements had been identified in modern scholarship as defining characteristics of the Chan tradition, including some that were not, I found, actually invented or used exclusively by monks traditionally identifed as members of the Chan Lineage.

As I studied the nominally Chan monasteries that flourished in Song and Yuan China, moreover, I came to realize that those instititions were scarcely different in their organization and operation from other large public monasteries that had no association with the Chan Lineage, and that all of them had evolved directly from mainstream Buddhist monasteries in the Tang. The so-called “rules of purity” (qinggui 清規) that regulated them were not the invention of Baizhang or the Chan movement in general, but belonged rather to what may be loosely called the Chinese vinaya tradition, in which Indian vinaya texts were translated, collated, interpreted, reformulated, and augmented over the centuries with indigenous Chinese rules, procedural guidelines, ritual manuals, calendars, and liturgical texts.

Now, after many years of research in this area, I have come understand the etiology of the ambiguities and inconsistencies in the uses of the words Chan and Zen that I initially confronted years ago. Back when I was a graduate student, however, my dawning awareness of those problems threw me into confusion. When I began to study the historical development of the “Zen monastic institution” in China, I thought I knew what that was, but the more I learned the less certain I became that such a thing had ever existed. But if it had not existed, then what was I researching? Was my topic now the history of a myth — the Baizhang story — that evidently had its roots in tenth-century China but had also gotten a big boost and a veneer of scholarly respectability in twentieth-century Japan? I had no doubt that there was in modern Japan a set of institutions (training monasteries, ordinary temples, universities, research institutes, publishing houses, etc.) that could meaningfully be grouped together under the rubric “Zen School,” and that the history of said school could be traced back to the Kamakura period. But, having realized that the Chan School in medieval China never had an independent institutional identity similar to that of the Zen School in Japan, I was confronted with a big question: what was the Chinese “Chan School,” anyway?


Types of Definition

Concerns such as these led me to address the problem of defining Chan as an object of historical study in my doctoral dissertation, where I attempted to frame the discussion in the context of what Western logicians have said about the process of definition itself.[17] In particular, I drew on a typology of definition articulated by the philosopher Richard Robinson in a book entitled Definition,[18] using that to analyze the different ways in which modern scholars have framed the Chanzong 禪宗 — variously translateable as the Chan “lineage,” “school,” “sect,” “movement,” or “tradition” — as an object of study. I summarize Robinson’s typology here because the distinction he draws between “lexical” and “stipulative” definitions informs the approach that I take in this book to the problem of defining Chan.[19]

Robinson begins by explaining the difference between “real definition” and “nominal definition.” A real definition is one that assumes the independent existence of the definiendum, understood as a real thing (L. res). Real definition of an object is always an analysis of that object. It is often an attempt, in the manner of Aristotle, to describe “the essence of the thing,”[20] or to identify some property that all the individual members of a genus hold in common (i.e. that which is “essential” to the genus). A nominal definition, on the other hand, is one that has as its definiendum not things as such, but words or concepts (L. nomina). The aim of nominal definition, Robinson says, is “to report or establish the meaning of a symbol.”[21]

Robinson acknowledges that many eminent philosophers, from Plato and Aristotle to Spinoza, Kant, and Dewey, have believed in real definition, but he argues that it is a spurious category and that all valid definition is actually nominal. Real definition, which searches for an essence, is rejected by Robinson on the grounds that “there is no such thing as essence in the sense intended.”[22] “Essence,” he argues, “is just the human choice of what to mean by a name, misinterpreted as being a metaphysical reality.”[23] Whenever people engage in intellectual activities they believe to be real definition, he maintains, they are either engaged in logically flawed procedures that should be abandoned altogether or are involved in legitimate and useful procedures (such as nominal definition) that they wrongly conceive as real definition.[24]

If, for example,[25] one thinks that the way to define “mouse” is to pinpoint the essential properties that all mice hold in common, one will probably proceed by doing something that is actually quite different: looking for a shared meaning, a common denominator as it were, in all the applications of the word “mouse.” But that is a fool’s mission, for (1) a rodent with a pointed snout and long tail, (2) a timid and quiet person, and (3) a device used to move a cursor on a computer screen do not necessarily have any properties in common that inspire us to give them the same name. The meanings of words get extended in all sorts of metaphorical ways that stray far from their original signification. Thus, for instance, there is now a wireless mouse that looks and functions like the older type of mouse that connects to a computer with a wire, but no longer bears much resemblance to the long-tailed rodent.

Within the category of nominal definition, Robinson distinguishes two types. The first, called “word-word definition,” reports or establishes the meaning of a symbol by saying that it means the same thing as another symbol. The second type of nominal definition, called “word-thing definition,” reports or establishes the meaning of a symbol by correlating it with a certain thing. I argue below that the process of correlating words with things may in some cases involve the initial delineation of the thing named; that is to say, the “thing” in question does not necessarily exist as such before it is defined.

Nominal definition, in Robinson’s scheme, may also be divided into “lexical definition” and “stipulative definition.” Lexical definitions report the meanings that people actually assign, or have assigned in the past, to a symbol within a particular linguistic or cultural milieu. Because lexical definitions attempt to establish the facts of actual usage, they need to be backed up by evidence (i.e., by citing examples of usage in contexts where the meaning is clear), and they may be judged as either true (attested) or false (unattested). Stipulative definitions, on the other hand, function to establish the meanings of symbols for future use within a particular area of discourse. They proceed by coining a new word that is correlated with some newly invented or previously unrecognized thing or phenomenon, or by taking an existing word and restricting its use to a single referent in order to purge it of all ambiguity. Stipulative definitions, because they set terms at the beginning of a study or debate, are essentially arbitrary: they cannot, in principle, be judged true or false on the basis of evidence of any sort. They can, however, be judged useful or not, and other people are free to adopt or ignore them. A stipulative definition that gains widespread acceptance, of course, thereby becomes a lexical one.

To illustrate the difference between lexical and stipulative definitions, let us consider the word “city.” Many things have been called cities, including: densely populated metropolitan areas; the older centers of such metropolitan areas; towns of any size that are incorporated; very small towns that have the words “City” in their name; necropolises or “cities of the dead” that have many graves but no living residents; “prairie dog cities” where large numbers of ground squirrels live in interconnected burrows; the “Celestial City” where God and the beatified are said to dwell; and so on. The word city has all of those attested meanings, which are thus part of its lexical definition. When professional demographers and urban planners study cities, however, they typically use a stipulative definition of “city” that specifies a minimum population (or population density) of human beings living within a certain area. That is an exercise in disambiguation, and one that is necessary at the outset of any scientific research on cities.

Imagine, if you will, an argument between a person who lives in Canyon City, Oregon (population 703 in 2015) and a demographer, with the former asserting that his place of residence is a city and the latter maintaining that it is not. To the extent that they both think they are debating what the place “really is” — what Robinson calls a matter of real definition — they are both in error. What they are actually arguing about is the application of the word “city” — a matter of nominal definition — with the demographer positing a stipulative definition (which is his prerogative) and the resident countering with a lexical definition that, as a matter of historical usage, is also correct. Since both parties arrive at their conclusions using justifiable criteria and sound reasoning, and each thinks the other is out of touch with some objective reality, such arguments can become quite heated. Many fights that rage over what “is” or “is not” existing in the real world can be shown, when analysed in this way, to simply be disagreements over the usage of words.


The Doctrine of Emptiness and its Implications for Definition

Although the terminology I employ is borrowed from Robinson, my thinking on the question of definition has been deeply informed by the Mahāyāna Buddhist doctrine of emptiness (kong 空, S. śūnyatā). In particular, when considering what Robinson calls “word-thing definition,” that doctrine has led me to reflect on whether and in what sense “things” as such can be said to exist before they are named. The discussion of emptiness that I offer in the following paragraphs has two purposes: to explain the underpinnings of my own views on matters of definition in historical inquiry, and to introduce a mode of philosophical analysis that was embraced by many adherents of the Chan School in medieval China.

The concept of a “thing,” if we analyze just what we have in mind when we use that word, is similar in many respects to the concept of a dharma (fa 法) as that was formulated in Buddhist Abhidharma literature. A dharma is a defined as a discrete entity that cannot be analyzed into parts; it has “own-being” (zixing 自性, S. svabhāva), which means that it exists independently and has an intrinsic nature — also called an identifying mark (xiang 相, S. lakṣaṇa) — that is unvarying. Nevertheless, dharmas are said to be impermanent (they exist for just an instant) and to arise and cease in a causal nexus. A “thing” as we ordinarily conceive it can, on the contrary, be made up of constituent parts, but it is similar to a dharma in that conceptually it is always one thing: a single, indivisible entity. We call a wristwatch a thing, for example, but we do not call two watches a thing, nor do we call a pile of watch parts a thing, unless we know that they come from a single timepiece that has been disassembled, or the thing we have in mind is that single pile.

Moreover, a “thing” as we conceive it has clear boundaries that separate it from other things, and it has identifying characteristics that are static. A “thing” is always just what it is; if it changes, then that is some other thing. An apple, when conceived as a thing, is always an apple; it is never soil. The observable fact that apples fallen on the ground rot and become soil does not keep us from thinking that those are two different things, albeit ones that are related to one another. We conceive of “things” as coming into and going out of existence, and as being causally interrelated with countless other things.

The Mahāyāna doctrine of emptiness, as expounded by thinkers such as Nāgārjuna in texts belonging to the Mādhyamika tradition, holds that the concept of a “dharma,” as defined in the Abhidharma, is self-contradictory in a number of ways. Any dharma one could conceive of, for example, must have some kind of imagined temporal duration and spatial extension that renders it theoretically susceptible to analysis into parts (a beginning, middle and end; a top, bottom, left and right). By definition, however, nothing that is made up of parts has “own-being.” A dharma that is conceived as arising in causal dependence on other dharmas, moreover, cannot also be considered an independently existing entity, i.e. one with “own-being.” Those are deficiencies demonstrated by logic, but Mādhyamika texts also make an empirical argument: when one goes looking for dharmas, none are ever found, because nothing in the real world meets the definition. The very idea of a dharma is thus an “empty” category, which is what the texts mean when they speak of the “emptiness of all dharmas” (yiqie fa kong 一切法空, S. sarvadharma-śūnyatā).

A similar critique may be applied to the idea of a “thing.” As explained above, we conceive of things as discrete, self-contained entities that have distinguishing characteristics and boundaries that separate them from other things. But “thing,” so defined, is an empty category or null set:[26] nothing in the real world actually has the properties that we associate with being a thing. All the “things” we identify are in reality composite, they are bound up in an infinitely complex web of causes and conditions, and they lack the clear-cut borders and duration without change that we ascribe to them conceptually.

Consider, for example, rivers such as the Hudson or the Mississippi, which in ordinary parlance are considered really existing things. Rivers, as we imagine them, have boundaries that separate them from other things around them: the earthen banks on either side, the sky above, the oceans into which they flow, and so forth. Moreover, rivers have identifying characteristics (e.g. a natural flow of water through an established channel of a certain size) that remain fixed and distinguish them from other things. But if we look more carefully at the dynamic process we call a river, it demonstrably lacks the properties of “thingness” that we impute to it. Upstream, where rain falls over thousands of square miles and water flows down in countless rivulets, brooks and tributary streams, just where are the dividing lines between those things that are not rivers and the thing that is a river? Downstream, as the water flows into a delta or estuary that opens onto the sea, just where does the river end and the ocean begin? Such boundaries exist in our imagination when we think of a river as a thing, but no such demarcations are found in the real world. There is just a continuous process of water evaporating, forming clouds, falling to earth, and flowing down to the sea. It is we who, in our collective (linguistically determined) and individual imaginations, draw the lines that divide that process up into the separate “things” that we call rivers, oceans, clouds, rain, and streams. Such imaginary lines, morever, can be drawn anywhere for any reason: there is nothing in the real world that stops us from delineating and naming whatever things we please, however useful or impractical the designations may prove to be.

In making this observation, I do not mean to deny the obvious fact that human cognition operates on pre-verbal and non-verbal as a well as verbal levels, and that we are hardwired to see, hear, smell, taste, or feel — and react to — many “things” in the world, whether or not we have attached names to them. Animals with large brains evidently experience the world in much the same way as humans do — as a place filled with discrete entitities that have distinctive appearances, movements, smells, flavors, etc. — and they do so without linguistic abilities that come anywhere near those of Homo sapiens. The vocabularies of early human languages, whenever our distant ancestors began to develop those, no doubt began with the naming of things in the environment that were already familiar and understood to be important. Word-thing definition, I readily concede, is rooted in modes of cognition that are not dependent on language, which is why we can teach a dog to correlate the sounds of the spoken words “ball” or “stick” with objects that the animal has already discriminated through non-verbal processes.

Word-thing definition, however, also includes the naming of things that do not immediately impinge on any of our senses, or are purely conceptual entities, and thus only present themselves as existent at the point when we name or conceive of them. Constellations, for example, do not exist as such until they are defined; a dog might be able to see the stars that make them up, but it has insufficient language skills to ever see Orion or the Big Dipper.

Regardless of how they are cognized, through the medium of language or through processes that do not depend on language, the fact remains that “things” as such do not exist. This does not mean that nothing exists, only that existence, in all of its dynamic and causally interrelated complexity, does not actually take the form of discrete, static entities. In this sense, there are no rivers, oceans, clouds, rain, or streams — no “things” of any sort. Nevertheless, there is some kind of really existing stuff, registering somehow in sense data and behaving in a manner that is regular and largely beyond our control, which is “there” (or “here”) whether or not we selectively cognize, discriminate, reify, and name it as the real things of this world. This understanding, expressed in the plainest English I can muster, is my interpretation of what philosophers of the “middle way” (S. Mādhyamika) meant when they claimed to avoid the extremes of asserting existence (you 有, S. bhāva) on the one hand and nonexistence (wu 無, S. abhāva) on the other. To rephrase their position in the terms that I have just laid out: it is true that no things exist, but it is not true that nothing exists.

The fundamental problem is that language, and all discursive thought that uses the medium of language, is utterly dependent on the naming of dharmas or “things,” but that very act of mental discrimination (fenbie 分別, S. vikalpa) distorts reality by drawing lines where none actually exist and by freezing dynamic processes into quasi-static entities. To point out that language necessarily employs distorting reifications, however, is not to argue that it is dysfunctional. On the contrary, language has proven such an effective and powerful tool for adapting to and manipulating our environment that we Homo sapiens, as a species, seem to have bet all our marbles on its evolutionary development.

We pay a stiff biological price for the huge brains that catalogue and correlate all of the concrete and abstract “things” of our world, past, present, and future: the size of our crania expose human mothers to a high risk of mortality in childbirth; human offspring need many years of development outside the womb before they can begin to fend for themselves; and our brains burn lots of calories, raising our requirements for nourishment. The psychological price we pay for language is high, too, for purely imaginary phenomena such as “what I will experience after I die” or “the meaning of life” can be triggers for high anxiety and deep depression, and we suffer greatly when the stories we formulate concerning our own identities and life goals fail to unfold as hoped or expected.

One of the functions of religious narratives, it would seem, is to mitigate the anxieties and disappointments that language, for all of its evolutionary advantages, exposes us humans to. Thus, for example, the gratuitous distress that may arise when contemplating “what I will experience after I die” can be neutralized with an equally imaginary story about a blissful afterlife in paradise. Similarly, all unhappy events in the narrative of an individual’s life can be reframed as but twists in a plot whose dénouement takes the form of a revelation of the Creator’s ultimate purpose, which is always to the good.

The Mahāyāna doctrine of emptiness — a species of religious narrative — is cleverly designed to admit the usefulness (and indeed the indispensability) of language, while at the same time inoculating us against the pathogenic belief that the “things” of this world actually exist as named, in and of themselves. From the point of view of the emptiness doctrine, neither the thing called “my self” nor the event called “death” have any ultimately real referents, so it is foolish to put much stock in tales that revolve around those narrative elements, regardless of whether the stories are frightening or reassuring. On the other hand, when it comes to the kind of conventional wisdom that deems it a good idea to write a will stipulating the disposition of one’s property after one’s death, there is nothing in the emptiness critique that prevents a common-sense adherence to that practice.

If language is dependent on the naming of dharmas or things, and no such things actually exist in the way that we imagine, it follows that so-called reality is beyond any perfectly accurate or ultimately true description. The real world of “things as they are” — referred to in Mahāyāna texts as “suchness” (zhenru 眞如, S. tathatā) — is in plain view, not hidden behind any sort of mystical veil that must be dramatically torn away if we are to behold it face to face, but there is no way to conceive or speak of it without immediately involving ourselves in the error of positing “things.” This, as I understand it, is what the Mahāyāna philosophers mean when they speak of the ineffability or inconceivability (buke siyi 不可思議, S. acintya) of the ultimately real.

At the level of conventional designation, nevertheless, the modes of discriminating and naming things employed in ordinary language do give us the ability to formulate more or less reliable and readily communicable stories about the world we live in. Language enables us to imagine, predict, and plan for the future; to reason based on past experience and formulate hypotheses for future testing; to shape our communications in ways that can influence the behavior of others, up to and including including the manipulation of others through the use of deliberately false statements; and to tell stories that, while admittedly “made up,” are nonetheless edifying or amusing.

The medium of language is intrinsically devoid of any sure connection with the real world, but it is undeniably powerful and fraught with very real consequences. That is why all languages have ways of differentiating modes of discourse (or genres of writing) that are “factual” versus those that are “fictional,” “actual” versus “hypothetical,” “firsthand” versus “second hand,” “about the past” versus “about the future,” and so on, as well as grammatical devices for signalling which mode a speaker is employing at any given time. All languages also draw basic distinctions between statements that are “accurate” or “inaccurate” as descriptions of some objectively verifiable state of affairs, and “true” or “false” in the sense of being reliable or unreliable as a basis for action. All systems of ethics, moreover, distinguish “true” and “false” statements at the level of human intention, placing a high premium on truthfulness (communicating in good faith as opposed to lying) with regard to both the mode of discourse a speaker claims to be using and the contents of what he or she says.

Mādhyamika proponents of the doctrine of emptiness, while they held that all verbal formulations predicated on the existence of “things” are ultimately false, were not so naïve or impractical as to deny the relative truth or falsehood (reliability or unreliability) of statements made in ordinary speech. An important corollary of the teaching of emptiness, already alluded to at several points above, is the doctrine of two truths (erdi 二諦, S. satya-dvaya): conventional truth (sudi 俗諦 or shidi 世諦, S. saṃvṛti-satya) and ultimate truth (zhendi 眞諦 or diyi di 第一諦, S. paramārtha-satya).

On the level of conventional truth, which is to say, within the conventionally agreed upon parameters of the lexicon and grammatical rules of any given language, statements about all the things of this world and the ways in which they interact can be judged either true or false on the bases of empirical evidence, adherence to rules of grammar and logic, internal consistency, accordance with other statements held to be true, the perceived reliability of the speaker, and so on. On the level of ultimate truth, however, all statements about “things” (whether conventionally true or not) are false. Sayings such as “all dharmas are empty” or “fundamentally there is not a single thing” (benlai wu yi wu 本來無一物) are meant to point to the ultimate truth, but they are only true on the conventional level because, in the final analyis, there are no such things as dharmas about which one can predicate emptiness, and no such thing as emptiness, either. A grasp of ultimate truth depends on conventional truth, for the former is simply the realization that language and conceptual thought can never get a handle on the real without distorting it in the process. Ultimate truth, in other words, is not some “thing” that can be intuitively grasped or experienced in and of itself, prior to or apart from language; it is just a name for an insight into the inherent limitations and delusive (albeit useful) properties of language.


The Need for Stipulative Definition in Historiography

Given the epistemological principles that I have outlined above, let us now reflect on the nature of historical inquiry in general, and the question of how to define Chan as an object of historical research, in particular. The things that historians research and write about are typically complex events and phenomena in the realm of human activity, such as social and economic institutions (e.g. empires, nation states, slavery, capitalism, communism), political movements (e.g. wars, revolutions, struggles for human rights), cultural developments (e.g. agriculture, technology, art, architecture), belief systems (e.g. philosophies, religions), and so on. Such objects of study are amorphous and highly abstract: as “things” go, they present themselves to us rather in the manner of  constellations that must be named to be seen, not in the way that a dog smells, sees, and bites a bone.

Historians, therefore, are in principle obligated to designate the identifying marks and boundaries of the events or phenomena they propose to investigate. The event in question needs to have at least a chronological starting point, prior to which the historian may speak of its “antecedents” and “causes” but not the thing itself. Often the end point of an event is marked as well, in which case its “results” and “repercussions” may be assessed. During the period of time in which the main event or phenomenon is said to have been actively unfolding, moreover, the historian must have some criteria for determining which lesser occurrences, among all those taking place simultaneously in the universe, were part of the main thing, and which were ancillary to it or completely irrelevant.

The Allied invasion of France on D-Day in 1944, for example, is always treated as an integral part of World War II. The weather in the English channel on D-Day had some impact on the planning and outcome of the invasion, so it is included as relevant to the history of the war. The birds that flew over the French beaches on D-Day, on the other hand, had nothing to do with the war and are not part of that event… unless, of course, a historian discovers some incident (e.g. warplanes that crashed after hitting birds) that connects them directly to the story. In short, it is the work of historians, individually and collectively, to selectively draw the lines that bring the “thing” — the event or phenomenon that is under investigation — into focus and distinguish its history from the history of everything else in the world, which is infinite in scope and complexity and thus beyond telling.

This is not to say, of course, that historians are free to “make things up” out of whole cloth, in the manner of novelists whose main sources of information for the stories they tell are their own imaginations. Historiography is, by conventional agreement, a genre of writing that is factual, not fictional. It speaks mainly of things that are believed to have actually happened in the past, although the historian is also given license to engage in a certain amount of hypothetical and suppositional thinking: to speculate on what “might have happened” if events had unfolded somewhat differently, and to draw reasoned conclusions about what “must have happened” in light of known facts, even in the absence of direct evidence.

Historical narratives, with the exception of autobiographies and personal memoirs that deal with recent events, generally take the form of second-hand rather than eye-witness reporting. That is to say, the historian begins with stories that other people have told and data that others have recorded in the past, most often preserved in the form of written documents, then selects and edits them for relevance to the historical event or phenomenon that is under investigation and weaves them together into an overarching narrative of what happened. Key to that process, of course, is the vetting of source materials for their accuracy and reliability, which involves judging such things as the scientific possibility of the events recounted, the trustworthiness and biases of the witnesses, the existence of corroborating accounts, physical evidence that may be found in artifacts or archeological remains, and so on.

Historians also divide their sources into “primary” materials that date from around the time of the event or phenomenon in question, and “secondary” materials, which are the recently published findings of other historians in the same field. Virtually all modern historiography builds on, refines, and/or takes issue with an existing set of historical narratives found in relevant secondary sources. At this level of discourse, historians compete with each other to produce the most compelling arrangements of factual data and interpretations of events, and the aims and methods of historiography itself get called into question in debates that become increasingly theoretical and abstract.

Good historiographical practice, in any case, calls for an explicit definition of the object of study at the outset of research, or at least at the start of a report on one’s research findings. It is true that some relatively recent and hugely disruptive “events,” such as World War II, are so deeply inscribed in the collective consciousness of large numbers of people, and have been worked over by so many historians, that a stipulative definition may not seem necessary. In such cases, perhaps, a historian can rely on the ordinary lexical definition of the thing in question — a conception of the thing that is relatively unambiguous and common knowledge — and just start talking about it without any risk of misunderstanding.

When it comes to historiography that deals with “Chan,” however, no clear consensus exists among modern scholars as to the parameters of the thing under discussion. The Chinese texts that scholars today rely on as primary sources for the “history of Chan,” moreover, use the word chan in a manner that is highly ambiguous, referring in some to cases to things (e.g., the practice of seated meditation) that we would call real, and in other cases to phenomena (e.g., demonstrably fictional lineages of Indian patriarchs) that we would label mythological. Under these circumstances, it is imperative that historians of Chan explicitly stipulate their object of study. What Robinson calls “real definition,” the misguided approach that assumes the independent existence of the definiendum and sets out to determine its essential qualities, inevitably results in research findings that are more or less confused and inconsistent.


Basic Categories and Definitions Employed in this Book


“Lexical History”

By “lexical history,” a term that I use in the title of Part One of this book, I mean historical research that seeks to determine the meanings that people have given to particular words or phrases in the past, as attested in whatever contemporary writings or other media survive from the places and periods of time one wishes to investigate. Being primarily concerned with the uses and meanings of words, lexical history does not inquire into the existence or true nature of the entities or phenomena that the words in question ostensibly refer to. Nor does it assess the “accuracy” of the definitions it encounters, in the sense of gauging the degree of correspondence between words and what are assumed to be extra-lingual realities. The only truth that a lexical history seeks to determine, and the only judgement that it makes, is whether a given word can accurately be said to have had a particular use or meaning in the past. The primary concern of lexical history, in short, is to determine the meanings of certain words within specific linguistic contexts, taking into account such things as grammar and syntax, semantic frames of reference, figures of speech, modes of rhetoric, and literary genre.

It is clear, however, that words also have broader social, political, cultural, and intellectual contexts that must also be taken into account if one is to fully appreciate their meanings at a given time and place. Thus, although it is not the primary aim of lexical history to determine “what happened” (apart from the use of words) in the past, it cannot ignore the outlook on and knowledge of the world possessed by the authors of the texts it takes as primary sources, and cannot avoid hazarding some estimations about the intentions that informed their writing. Lexical history can be distinguished from other modes of historical investigation that focus on events, circumstances, and protagonists, and that attempt to establish chronological and causal relationships between them. In the final analysis, however, it can only operate in conjunction with those other modes of historical inquiry.

The lexical history of the word “chan” 禪 that I present in Part One of this book is a study of the evolution of the meanings of that word, in all combinations and contexts, from the time of its earliest appearance in Chinese Buddhist texts down through the Yuan dynasty. In particular, I detail how the meanings of the word chan evolved in Chinese Buddhism, from (1) its early use as a rendering of the Sanskrit dhyāna or “meditation” to (2) its redefinition in the seventh and eighth centuries as a synonym of prajñā or wisdom (hui 慧), (3) its use from the late eighth century as a name for various putative lineages of ancestral teachers (zushi 祖師) in which an Indian monk named Bodhidharma (Putidamo 菩提達摩 or 菩提達磨) was posited as the founding patriarch, and (4) its final emergence in the Song dynasty as the name of a school of Buddhism that had a sizable living membership and a vast family tree of spiritual ancestors. By the time that the word chan came to used in the name of the “Chan Lineage” (Chanzong 禪宗), its meaning had changed so much that it no longer meant simply “meditation,” but was also a synonym for awakening (puti 菩提, Skt. bodhi) or the “mind of buddha” (foxin 佛心). Thus, in Song China, another name for the Chan Lineage was “Buddha-Mind Lineage” (Foxinzong 佛心宗). All those extensions of the meaning of chan were incremental: none of the earlier meanings were lost when new meanings were added.

Basic philological research of this type is a prerequisite to any study of the history of Chan (however that phenomenon is defined) because the word chan is highly ambiguous, and scholars cannot assume that whenever it occurs in a documentary source it necessarily pertains to their chosen object of study. In China, the number of binomial and polynomial words containing the glyph chan 禪 increased over the centuries as new terms were coined, and older words containing that glyph also took on new meanings. As a result, modern scholars are at considerable risk of interpreting occurrences of the glyph chan in ways that are anachronistic or otherwise inaccurate.

Modern scholars have not been entirely oblivious to the ambiguities of the word chan, and a few have devoted articles or sections of books to discussions of its meanings in the contexts of various Chinese Buddhist texts. Sekiguchi Shindai 關口眞大, for example, wrote a pioneering essay entitled Zenshū no hassei 禪宗の發生 (“The Genesis of the Chan Lineage/School”) in which he tried to determine exactly when the expression “Chan Lineage” (chanzong 禪宗) began to be used in reference to followers of Bodhidharma.[27] Yanagida Seizan 柳田聖山, responding to Sekiguchi, addressed the same question in his Shoki zenshū shisho no kenkyū 初期禪宗史書の研究 (“A Study of Historical Documents Pertaining to the Early Chan Lineage/School”).[28] However, Part One of the present book is the first attempt ever to present a thorough-going lexical history of the word chan.

As a general rule in this book, whenever I determine that a particular instance of the word chan in a Chinese text is used clearly and unambiguously as an equivalent of the Sanskrit word dhyāna, to translate it into English I render it either as “meditation” or as “dhyāna,” which is the anglicized form of “dhyāna.” Whenever I decide that the word chan functions clearly and unambiguously as a proper noun or adjective to refer to a lineage or teaching associated with Bodhidharma, I anglicize it as “Chan,” capitalized. Finally, in cases where the word chan is used in a way that is unclear or ambiguous, I decline to translate it, leaving it in romanized Chinese, which is by convention written in lower case italics. I also use the romanized Chinese, of course, when introducing the word chan as a topic of discussion, as for example in the opening sentence of the present paragraph.


“Chan Lineage” vs. “Chan School”

For the purposes of this book, I reserve the English word “lineage” for use in translating what the traditional histories call a zong 宗: a genealogy comprised of ancestral spirits and (in the most recent generation or two) their living heirs. When I speak of a historical “school” of Buddhism, on the other hand, I mean a group made up entirely of flesh-and-blood, living people who were united by some set of beliefs, practices, or social arrangements that they held in common, and whose circumstances and activities are in principle open to historical investigation.

In accordance with these working definitions, a truly critical history of the Chan Lineage in medieval China would treat it as a mythological entity, the construction of which can be traced diachronically through a sequence of Tang and Song texts. It would not assume the existence of a real social entity — what I call a school — that necessarily had a membership similar to that of the Chan Lineage as it is depicted in those texts. Rather, it would seek to identify the concrete social and institutional contexts in which the myth of the Chan Lineage (the Lineage of Bodhidharma) took shape; research the ways in which the myth served the ideological needs and political interests of the various parties who promoted it; and investigate what influence the myth had (e.g. through ritual reenactment and the privileging of certain individuals) on concrete social structures and modes of religious practice. Modern scholars have sometimes taken such approaches in limited contexts, but I am the first to draw a clear and explicit distinction between the Chan Lineage as a product of the religious imagination and the Chan School as a collection of real people who did the imagining.[29]

Part of the problem is that, in modern Japanese and Chinese usage, the words shū 宗 and zong 宗 can refer both to real social entities, such as the Sōtō School (Sōtōshū 曹洞宗) of Zen that is registered with the Japanese government today as a non-profit religious corporation, and to entirely mythological entities, such as the Chan Lineage (Chanzong 禪宗) in India, which according to the traditional histories comprised twenty-eight generations of ancestors culminating in Bodhidharma. It may be clear from the context whether the shū 宗 or zong 宗 in question is imaginary or real (a “lineage” or a “school”), but in too many cases the modern scholarship remains ambiguous on this matter. When modern historians speak of the Chanzong 禪宗 in Song China, for example, are they talking about the relatively small number of eminent monks who are named in traditional histories as dharma heirs in a Chan Lineage believed to “transmit mind by means of mind” (yixin chuanxin 以心傳心), or about a Chan School of Buddhism that had tens of thousands of clerical and lay adherents and dominated the upper echelons of the state-supported Buddhist monastic institution? In most works of modern Japanese and Chinese scholarship, this question is neither asked nor answered. When the object of study — the Chanzong 禪宗 — remains so poorly defined, the research findings are bound to be muddled.

This problem is perhaps more apparent to translators who need to render the Sino-Japanese word zong 宗 (J. shū) into a language that does not use Chinese characters than it is to native speakers of Japanese or Chinese working in their own languages. After all, the ambiguity that I impute is rooted in the modern (Western) distinction between “mythological” and “historical” entities. Although scholars in East Asia today certainly understand and make use of that distinction, it was far from anyone’s mind in the early ninth century when the term Chanzong 禪宗 first entered the vocabulary of Chinese Buddhists as a name for the Lineage of Bodhidharma, and old linguistic habits are hard to break.

For the purposes of this book, I define the Chan School of Buddhism in medieval China as the group of people — monks, nuns, and lay followers — who were united by a shared belief in a multi-branched Chan Lineage, the latter being conceived as an extended spiritual clan that was founded in India by the Buddha Śākyamuni and transmitted to China by a first ancestor (chuzu 初祖) named Bodhidharma. That concept of a Chan Lineage does appear in a few texts dating from the early ninth century, but it was not widely known at that time and did not gain significant credence among Chinese Buddhists until the middle of the tenth century.

The Chan School, as I define it, thus seems to have come into its own in the decades preceding the advent of the Northern Song dynasty in 960. Over the course of the next century, it developed into the most prestigious and powerful tendency within the state-approved Buddhist monastic institution and remained so throughout the Southern Song and Yuan dynasties. Most monk and lay followers of the Chan School in those times were not, by the school’s own standards, recognized as members of (i.e. dharma heirs in) the Chan Lineage, although a few monks did attain that lofty goal, which was called “attaining buddhahood and becoming an ancestor” (chengfo zuozu 成佛作祖).

The definition of the Chan School that I stipulate here is bound to be controversial, for almost all modern historians regard the Chan School as an entity that developed much earlier in China, if not with Bodhidharma himself, then at least with one or more of the monkish sodalities that claimed descent from him in the eighth and ninth centuries. However, the modern scholarship has not reached any concensus on the most basic question of definition, which is: what ideas (doctrines, beliefs, etc.), practices, or modes of social organization should be taken as the identifying marks and delimiting characteristics of the “Chan School”? As I explain in detail in Part Three, a number of different criteria have been proposed: the practice of meditation (chan 禪, S. dhyāna); the redefinition or outright rejection of meditation; the doctrine of sudden awakening (dunwu 頓悟); the rejection of Buddhist scholasticism; the rejection of Buddhist merit-making practices, especially rituals performed for the laity in exchange for material support and patronage; and the style of rhetoric that modern scholars refer to as “encounter dialogue” (J. kien mondō 機縁問答). In the final analysis, however, it is clear that most modern historians have not actually applied their own criteria of definition when delineating the membership of the Chan School in the Tang Dynasty, but have simply assumed the existence of a school that had the same membership as that given in traditional histories (genealogies) of the Chan Lineage. Thus, the question of how to define the Chan School has been left wide open in the modern field, and confusion reigns.

The definition that I offer here is a frankly stipulative one, conceived as a way to convert the traditional notion of membership in the Chan Lineage, which hinges on inheritance of an avowedly signless (wuxiang 無相) “mind dharma” (xinfa 心法) that is beyond the ken of any empirical testing, into a category that is suitable for use in critical historiography: the observable phenomena of stated belief in and ritual reenactment of a “Chan Lineage.” One benefit of my approach is that it maintains a clear distinction between those objects of historical study that are ideas (i.e. religious beliefs and myths associated with “Chan”), and those that are concrete phnemonena: the activities of living human beings who demonstrably voiced and accepted certain ideas and behaved in accordance with them. That may seem like a rather pedestrian standard for a work of critical historiography, but it is one that all too many modern studies of the history of Chan have failed to meet.

By defining the Chan School in a way that casts it as a product of the tenth century, I do not mean to say that it suddenly appeared out of nowhere, only that the set of beliefs that I take to be definitive of it came to be fully articulated and began to gain a widespread following at that time. That set of beliefs, which I call the “traditional history of the Chan Lineage,” evolved from earlier stories of a Lineage of Bodhidharma that had been elaborated over the course of the previous three centuries. Those stories and the texts (proto-histories) that helped to perpetuate them, moreover, were handed down within the Buddhist monastic institution. That institution underwent a period of severe official suppression during the Huichang 會昌 era (841-847) of the Tang and was disrupted by the political turmoil and wars that followed the collapse of the dynasty in 906, but it survived in various kingdoms during the Five Dynasties period (906-960) and reestablished itself with renewed vigor when China was again unified and relatively peaceful under the Song. The Chan School — the movement that produced, embraced, and acted out the traditional history of the Chan Lineage — spread rapidly within the context of that resurgent monastic institution and helped to gain supporters for it at the imperial court and among the broader class of elite scholar bureaucrats known as literati.


“Chan Lineage/School”

Whenever I am confronted with an ambiguous use of the terms Zenshū 禪宗 or Chanzong 禪宗 in the modern Japanese or Chinese scholarship that focuses on the history of that entity in China, I represent it in English with an equally ambiguous translation: “Chan Lineage/School.” This is necessary because, in my analyses of modern histories, I need a shorthand way of indicating the fact that they have taken a big step by employing methods of critical historiography learned from the West, but that they still have one foot caught in the conceptual framework that informs the traditional histories. The resulting intellectual posture is awkward and off balance, and my ungainly straddling translation — “lineage/school” —reflects that situation.


Three Classes of Historiography

Parts Two, Three, and Four of this book are organized around a set of fundamental distinctions that I draw between (1) traditional histories of the Chan Lineage, (2) modern histories of the Chan Lineage/School, and (3) proto-histories of the Chan Lineage. The main criteria that I use to place any given historical account (a text that exists today) within one of these three broad classes are: when and by whom the text was written; how (if at all) it conceives “Chan” as an object of inquiry; and what subsequent generations of historians (pre-modern as well as modern) made of it. The three classes of historiography and the specific criteria I use to delineate them are explained below.

At this juncture, let me stress that in distinguishing between traditional histories of the Chan Lineage, modern histories of the Chan Lineage/School, and proto-histories of the Chan Lineage, I reject any notion that they are simply three different kinds of accounts of one and the same historical phenomenon. The stories they tell are related to one another in complicated ways, but they are not stories about the “same thing.” Nor does the account of the various histories of Chan that I present in this book have exactly the same object of study as any of those histories, for what I engage in here is in large part a kind of “meta-history,” that is, a history of histories.

At the point when the modern study of the Chan Lineage/School first got started in late nineteenth-century Japan, virtually all of the documentary evidence that scholars had to work with belonged to what I call the class of “traditional histories.” The mark of a traditional history, as I define that term, is that the text in question posits or assumes a multi-branched Chan Lineage consisting of twenty-eight Indian ancestral teachers (the 28th being Bodhidharma), six Chinese ancestral teachers (the 1st being Bodhidharma and the 6th being Huineng), and a bifurcation in the generations following Huineng into two main branches that are still considered viable (as opposed to branches that have died out): the lineages of Mazu Daoyi 馬祖道一 and Shitou Xiqian 石頭希遷. Although the traditional histories present documentary records that ostensibly go all the way back to the founding of the Chan Lineage in ancient India at the time of Buddha Śākyamuni, the oldest extant work that meets the aforementioned criteria is the Ancestors Hall Collection (Zutang ji 祖堂集), first compiled in 952 in China, but lost there sometime in the early Song dynasty and surviving only in a revised Korean recension printed in the Koryŏ edition of the Buddhist canon in 1245.[30] The most influential traditional history, which provided the model for a number of later works that added subsequent generations of dharma heirs to the Chan Lineage, is the Jingde Era Record of the Transmission of the Flame (Jingde chuandeng lu 景徳傳燈録),[31] compiled in 1004.

I categorize several hundred texts dating from the Song and Yuan dynasties as traditional histories. Most of them belong to one of three genres: (1) records of the transmission of the flame (chuandenglu 傳燈録) or, as modern Japanese scholars have nicknamed them, “flame histories” (J. tōshi 燈史); (2) discourse records (yulu 語録) of individual Chan masters; and (3) kōan (gongan 公案) collections. In addition, there are a number of miscellaneous works that are not readily subsumed under any of these three classifications but may be regarded nevertheless as histories of the Chan Lineage as that entity has been traditionally conceived.

Among all the texts that I treat as traditional histories, the flame history genre takes precedence, for it is the body of literature that maps the basic contours of the Chan Lineage of ancestral teachers as that is traditionally conceived and explains the nature of the supreme dharma (fa 法) said to have been transmitted through that lineage from the Buddha Śākyamuni (and previous buddhas) down to the present. The archetypal text in this genre is the aforementioned Jingde Record, which was compiled in the Jingde Era (1004–1007) of the Northern Song Dynasty and provided the basis for many subsequent flame histories. The discourse records, kōan collections, and other texts that I categorize as traditional histories serve in various ways to augment and modify the conception of the Chan Lineage established in the flame histories, but they are all works of a derivative or secondary nature in that they take the genealogical records found in the flame histories for granted and elaborate on the words and deeds of monks whose identities as ancestors in the Chan Lineage have already been established in that literature.

The traditional histories depict Chan masters, especially in the generations following the sixth ancestor Huineng, engaging in a style of witty repartee or “question and answer” (wenda 問答) that may begin with a straightforward question concerning Buddhist doctrine or practice but quickly jumps to the level of indirect speech. They converse in colloquial (or quasi-colloquial) Chinese and use down-to-earth metaphors to express Buddhist teachings, avoiding the literary language, highly technical vocabulary, and syllogistic style that characterizes many translated sūtras and śāstras. The responses that masters give to their students’ questions often appear to be non sequiturs, and they are punctuated at times by equally enigmatic shouts, blows, and other dramatic gestures. Their words may be disconcerting: sometimes because they seem iconoclastic, antinomian, or sacrilegious, and sometimes because they appear to be utterly mundane or trivial. Nevertheless, it is assumed in the traditional histories that all heirs to the Chan Lineage speak and act with the immediate authority of the buddha-mind, so all of their sayings and doings, no matter how irrational (or merely pedestrian) they may seem at first glance, are framed as spontaneous and profound expressions of their awakening.

Early modern scholars had a critical bent that led them to evaluate the traditional accounts by means of comparative study, to distinguish “historical” from “mythological” elements in those accounts, and to seek additional evidence that might corroborate or disprove them, especially in the form of older texts that were presumed to be more reliable if they were closer in time to the events depicted. Nevertheless, the traditional histories provided the starting point for all such critical research, and the traditional vision of the Chan Lineage as a single family tree (albeit one with both living and dead branches) provided the basic conceptual model that modern scholars adopted without question when they first set out to write a more scientific history of the Chan Lineage/School.

Histories of the Chan (Zen) Lineage, of course, have been studied and written in Japan since the Kamakura period, so the question may well be raised: what substantive grounds are there for distinguishing between “traditional” accounts of the lineage produced from the thirteenth down through the nineteenth centuries and “modern” works produced by scholars in the field of Zen studies in the twentieth century? I make that distinction with some trepidation, for while the adjective “modern” is a word that remains vague and ambiguous in ordinary language, it is one that has been commandeered as a quasi–technical term in critical theory and thereby come to be loaded with all sorts of connotations (e.g. “modernity”) that I do not necessarily intend to invoke. Then too, “tradition” is a category that has attracted much critical rethinking, as scholars point out the rather obvious fact that many cultural phenomena that are ostensibly “received” from the past are demonstrably “invented” in the present moment (whenever that is) in which they are embraced. Be that as it may, my use of the categories “traditional” and “modern” in this book is decidedly low-tech and old-fashioned. I regard any history of Chan as “traditional” if it relies entirely on the concept of a lineage of dharma transmission from Śākyamuni to Bodhidharma and Huineng — a concept that has definitely been received from Song and Yuan dynasty China — as an organizing principle. On the other hand, I classify any history of Chan as “modern” if it calls the traditional conception of that lineage into question or attempts to redefine the Chan School of Buddhism in China in any other way. Those redefinitions are not received from traditional histories, but recently invented.

It is not my intention to argue that there was any kind of watershed that marked the absolute end of traditional Zen historiography in Japan and ushered in the modern field of Zen studies. The time–honored mode of historical writing that produced the Biographies of Monks in the Recent Zen Sangha (J. Kinsei zenrin sōbōden 近世禪林僧寶傳) in 1890, a work that I would classify as “traditional,” did not suddenly disappear at the turn of the twentieth century: it has continued, in one form or another, down to the present day.[32] Nor was the new style of “critical” historiography represented by works such as Sakaino Satoshi’s 境野哲 An Outline of the History of Chinese Buddhism (J. Shina bukkyōshi kō 支那佛教史鋼), published in 1907, entirely without precedent in earlier Japanese scholarship. I agree with Bernard Faure when he points out that Sakaino’s approach had a precedent of sorts in the Edo period scholar Tominaga Nakamoto 富永仲基 (1715-1746), who used comparative, text-critical methods to demonstrate the apochryphal nature of the traditional list of twenty-eight Indian ancestral teachers of the Chan Lineage and gave a “radically deflated image of Bodhidharma.”[33] What I mean by attaching the label “modern” to the works of Sakaino, Nukariya Kaiten 忽滑谷快天 (1867–1934), Okada Gihō 岡田宜法 (1882-1961), Yabuki Keiki 矢吹慶輝 (1879-1939), Tokiwa Daijō 常盤大定 (1870-1945), Ui Hakuju 宇井伯壽 (1882-1963), Suzuki Daisetsu 鈴木大拙 (1870-1966), and various other Japanese scholars who pioneered the field of Zen studies in the early twentieth century is not that they broke entirely with traditional modes of Zen historiography, but only that they were strongly influenced by methods of historical and textual criticism introduced from the West in the decades following the Meiji Restoration.

The nineteenth-century positivism that Japanese scholars learned from the West gave them a powerful new tool for reassessing the traditional histories, but it did not always result in a better understanding of those materials. There was a tendency in the early modern field of Zen studies to treat the sayings of Chan masters that are found in the traditional flame histories and discourse records not as religious literature, i.e. as sources for understanding how medieval Chan Buddhists understood the “awakened mind,” but as straightforward documentary evidence that could be used to extrapolate an accurate picture of the institutional settings in which they lived and the modes of Buddhist discipline and practice that they and their disciples were actually engaged in. The language of the “question and answer” (wenda 問答) genre, as noted above, is replete with apparently iconoclastic, antinomian, or sacrilegious sayings and gestures, ostensibly employed by masters as opportune devices (jiyuan 機縁) to bring their disciples to awakening. Reading that literature as if it contained verbatim records of things actually said and done, modern scholars envisioned a “golden age” of Chan in which the ancestral teachers of the Tang dynasty literally practiced what they preached, acting out their radical rhetoric in equally unconventional deeds.

By the middle of the twentieth century, the preponderance of modern scholarship had come to agree that the Chan Lineage/School originated as an iconoclastic, sectarian movement that rejected the mainstream Chinese Buddhist monastic institution, based as that was on an alien Indian model, together with all the “superstitious” beliefs and conventional modes of worship and practice that it embodied. According to that scenario, the followers of Bodhidharma in the first few generations of the Chan Lineage/School were all wandering ascetics. Chan monks did eventually settle down in their own monasteries, perhaps around the time of the fourth and fifth ancestors Daoxin 道信 (580-651) and Hongren 弘忍 (601-674), but those institutions existed independently of the rest of the Buddhist sangha. The Chan Lineage/School thus, purportedly, developed its own unique set of monastic procedures, which were codified by Baizhang (749-814), author of the first systematic “rules of purity” (qinggui 清規). Early Chan monasteries, as those are imagined in the modern scholarship, are said to have achieved economic independence through the practice of manual labor (chiefly farming) by the monks themselves, thereby avoiding the devotional prayers, merit-making rituals, and other rituals conventionally performed in exchange for lay patronage. Chan monks, it is further claimed, did not rely on Buddhist teachings handed down in sūtras or commentaries and did not get caught up in a lot of philosophical mumbo-jumbo. They sought instead an immediate, personal realization of awakening by means of meditation practiced in the midst of everyday activities, and by interacting with an accomplished spiritual guide — a Chan master, who generally served as abbot.

This scenario had a lot of appeal in Japanese intellectual circles when it was first introduced in the late nineteenth and early twentieth centuries, a time when traditional Buddhism in Japan was under attack for its “unscientific” world-view and “superstitious” concern with ancestral spirits. Early Chan, imagined as a refreshingly iconoclastic movement, seemed to hold out hope for a mode of Buddhist practice and spiritual attainment that was rational in its rejection of arcane religious dogma and ritual, accessible in principle to the laity as well as monks, and compatible with the task of turning the Japanese nation into a modern industrialized power. That idealized vision of a “pure Zen” (J. junsui Zen 純粋禪) also proved attractive in the West when it was introduced in the English language publications of the Japanese scholar D.T. Suzuki (1870–1966),[34] and it has now survived through several generations of modern scholarship that has been produced and consumed around the world.

The historical evidence for the existence of a Chan movement in medieval China that was literally (as opposed to just rhetorically) iconoclastic and sectarian, however, is weak, and the modern scholarship which made that case is rife with methodological errors. For example, when using traditional histories as sources for reconstructing the institutional arrangements of the early Chan Lineage/School, the scholarship failed to see that the realistic monastic settings in which the Chan ancestors of the Tang are depicted contain many anachronistic detail that betray the records as products of the Song. Nor did it occur to modern proponents of a “pure Zen” in the Tang that when a work of literature gives the reader access to the innermost thoughts of its protagonists, not to mention ostensibly verbatim records of conversations between master and disciple that are said to have taken place in the privacy of an abbot’s room, it is almost certainly a work of fiction. Furthermore, the modern scholarship has been prone to interpret the antinomian statements of Chan masters as straightforward recommendations for (or descriptions of) actual practice, rather than as rhetorical devices meant to emphasize the emptiness of dharmas — that is, to combat deluded attachment to the conceptual categories that are conventionally employed in Buddhist teachings. When iconoclastic statements are made in the context of Song Chan monastic institutions that were demonstrably conservative in their emphasis on proper decorum (weiyi 威儀) and embraced of every sort of conventional Buddhist practice, such statements can only be read as advice for the proper mental attitude with which to engage in those practices, not as documentary evidence that the practices in question were literally abandoned. When similar statements are put into the mouths of Chan masters said to have flourished in the Tang, however, modern scholars have felt free to take them literally.

As noted above, the modern field of Zen studies has never been willing or able to fully extricate itself, or even to gain much critical distance, from the basic conceptual framework established by the traditional (Song and later) histories. The Chan Lineage as understood in the traditional histories is a genealogy of ancestral teachers: monks who have attained awakening (wu 悟), inherited the dharma (sifa 嗣法) handed down from Śākyamuni through the line of Indian and Chinese ancestors, died and entered nirvana (rumie 入滅), and now exist in the form of awakened spirits (jueling 覺靈) who continue to interact with their descendants and others among the living by receiving offerings and answering prayers. This is a key set of religious beliefs that has informed the Chan School throughout its history, and it should be studied as such, but many works of modern scholarship have pursued research on the Chan Lineage as if it were a real social entity: a school of Buddhism whose members at any given time were all actual people, not ancestral spirits or any other kinds of fictive characters. Some scholars have attempted to verify and correct the traditional genealogical records by weeding out materials that are clearly mythological and corroborating the data that remains by using external sources, but that approach still treats the traditional concept of a spiritual lineage as if it were a viable category for use in critical historiography, which it is not.

The traditional histories, for all of their apparent naiveté, take a position that is more internally consistent than that of their modern counterparts. Those Song and Yuan works regard the Chan concept of “lineage”— a line of ancestral teachers through whom the formless, ineffable mind-dharma of the Buddha has been transmitted down to the present — as a historiographical category as good as any. From the standpoint of the Mahāyāna philosophy of emptiness that informs the traditional histories, however, all conceptual categories are ultimately fictive, albeit useful as linguistic conventions. Thus, the distinction that I labor to draw between schools and lineages as “real” versus “imaginary” entities, if it were raised as a topic for comment in the traditional “question and answer” (wenda 問答) literature for which Chan is famous, might warrant a trademark “shout” (he 喝) or “thirty blows” (sanshi bang 三十棒) for its foolishness. But the point of such chastisement is not that people should stop making distinctions or stop communicating altogether (the compilers of Chan records never did), only that we should not cling to any linguistically formed categories as if they corresponded perfectly to reality. Whenever we get caught in that self-imposed bind, we suffer when things do not go as imagined and hoped, and we lose sight of the fact that it is possible to make different and better — more refined and more effective — distinctions. The liberating realization expressed in the religious literature of Chan is that the real world is always more complex than we think.[35]

The designation “proto-history of the Chan Lineage,” as I define it, applies to any text that falls short of satisfying all the criteria for classification as a traditional history proper, but has been regarded by the compilers of the traditional histories or by modern scholars as a legitimate source for the history of the Chan Lineage, or as a work that prefigures the traditional histories of the Chan Lineage in one way or another. Thus, for example, any text dating from before 952 C.E. (the year when the Ancestors Hall Collection was compiled) that mentions Bodhidharma, whether or not it names him as the founder of a lineage (zong 宗), is a proto-history by my lights, for the compilers of the traditional histories and/or modern scholars of the Chan Lineage/School have used all such texts as historical sources. The same is true for all pre-Song texts that contain biographical information or teachings ascribed to any other monk who is named as an ancestral teacher in the traditional (Song) Chan genealogies.

Some of texts that I categorize as “proto-histories” of the Chan Lineage were known to and used by compilers of the traditional histories in the Song. Examples include: Platform Sūtra of the Sixth Ancestor (Liuzu tanjing 六祖壇經);[36] Baolin [Monastery] Biographies (Baolinzhuan 寶林傳),[37] a work in ten fascicles compiled in 801 by a monk named Zhiju 智炬 or Huiju 慧炬; and the works of the scholar monk Guifeng Zongmi 圭峰宗密 (780-841), such as his Chart of the Master-Disciple Succession of the Chan Gate that Transmits the Mind-Ground in China (Zhonghua chuanxindi chanmen shizi chengxi tu 中華傳心地禪門師資承襲圖),[38] written between 830 and 833, and his Preface to the Collected Writings on the Source of Chan (Chanyuan zhuquanji duxu 禪源諸詮集都序),[39] written around 833.

Other texts that I treat as proto-histories were either lost by the advent of the Song Dynasty or suppressed early in the Song, which meant that they were not handed down to posterity in the traditional histories. It is the modern field of Zen studies that has rediscovered this class of works, reclaimed them as relevant, and elevated them to high status as the oldest primary sources for the history of Chan. Modern Japanese scholars refer to the texts in question, many of which come from the cache of manuscripts discovered at Dunhuang in western China soon after the turn of the twentienth century, as “historical documents pertaining to the early Chan Lineage/School” (J. shoki Zenshū shiryō 初期禪宗史料). Noteworthy among the proto-histories in question are: records of the teachings of Heze Shenhui 荷澤神會 (670-762), such as the Treatise Determining the Truth About the Southern Lineage of Bodhidharma (Putidamo nanzong ding shifei lun 菩提達摩南宗定是非論);[40] the Record of the Transmission of the Dharma Treasure (Chuan fabao ji 傳法寶紀),[41] edited by Dufei 杜朏 sometime after 713; the Record of Masters and Disciples of the Laṅkāvatāra (Lengqie shizi ji 楞伽師資記),[42] written by Jingjue 淨覺 between 719 and 720; and the Record of the Successive Generations of the Dharma Treasure (Lidai fabao ji 歴代法寶記),[43] composed shortly after 775. All of these works contain some version of a lineage of dharma transmission said to have been founded in China by an Indian monk named Bodhidharma (or something similar), who figures prominently in the traditional (Song and later) histories as the first ancestor of the Chan Lineage.

A third group of works that I categorize as proto-histories are texts identified by modern scholars as sources for the history of the early Chan Lineage/School because they mention monks identified in the traditional histories as members of Bodhidharma’s lineage or contain teachings now attributed to them. Examples include: the so-called Long Scroll (J. chōkansu 長巻子) of Bodhidharma’s Treatise on the Two Entrances and Four Practices (Erru sixing lun二入四行論), first discovered by D.T. Suzuki in 1935 in the collection of Dunhuang manuscripts held at the National Library in Beijing;[44] the Great Master Bodhidharmas Treatise on Bloodlines (Damo Dashi xuemai lun 達磨大師血脈論);[45] Great Master Bodhidharmas Treatise on Awakening to Buddha-Nature (Damo Dashi wuxing lun 達磨大師悟性論);[46] Great Master Bodhidharmas Treatise on Breaking Free of Signification (Damo Dashi poxiang lun 達磨大師破相論);[47] and the Dharma Gate of Calming the Mind (Anxin famen 安心法門).[48]

Some of the proto-histories can be regarded as ideological precursors and literary prototypes of the traditional histories of the Chan Lineage, which took shape later. Others are sources that the traditional histories cited or borrowed from. I do not agree, however, with the modern scholarship that regards all such works as “historical documents pertaining to the early Chan Lineage/School.” For one thing, the proto-histories stand as evidence that, while various aspects of the traditional conception of the Chan Lineage were floated in the Tang, that conception did not assume its classic form or begin to gain widespread acceptance until the Five Dynasties period and Northern Song. Relatively few of the proto-histories actually refer to a “Chan Lineage” by name or show any sign of the traditional understanding of that lineage as a widely extended spiritual clan that has many legitimate branches. Moreover, because the proto-histories represent the views and interests of competing groups within the Buddhist order, they undermine the modern assumption that a single religious sodality worth calling the “early Chan School” existed prior to the tenth century.

To draw an analogy from the history of technology, it is clear that the first automobiles or “horseless carriages” were built by mounting steam, electric, or combustion engines on carriages similar to ones pulled by horses, but it would be erroneous (an example of the teleological fallacy in historical thinking) to study either the evolution of such carriages or the development of engines, which took place independently of each other in the preceding centuries, as the “history of the early automobile.” By the same token, while many of the myths, rituals, modes of rhetoric, and institutional arrangements that characterized the Chan School of the Song and Yuan dynasties had precedents or prototypes in the Chinese Buddhism of earlier periods, prior to the time that those elements all came together to form a nexus, it is misleading to speak of them as aspects of an “early Chan School.”

For these reasons, in this book I only turn to a detailed analysis of the proto-histories in Part Four, after dealing at length with the traditional histories in Part Two and modern histories in Part Three. In my view, a discussion of the traditional histories takes precedence because, without those records that were compiled in the mid-tenth century and later, which provided a conceptual lens and set of criteria for selection, modern scholars would never have identified any texts dating from the Tang or earlier as belonging to a “Chan” lineage, school, sect, movement, or tradition. Furthermore, it is only from the standpoint of modern scholarship that all of the proto-histories are relevant to the history of Chan: some are works that the Song compilers of the traditional histories either knew nothing of, or rejected as false and therefore unworthy of inclusion in official records of the Chan Lineage.

The class of traditional histories, although it includes several different genres of texts, is a relatively homogeneous one: it consists largely of biographies of Chan masters and sayings attributed to them. One reason for the homogeneity of the works in question, perhaps, is that the compilers of those histories were almost all monk or lay proponents of the Chan Lineage in the Song and later who wished to bolster and benefit from its prestige, not change the successful historiographic formula that had helped to create that prestige in the first place.

The class of proto-histories, by contrast, is more of a mixed bag, because it contains texts that prefigure the traditional histories in a number of different ways. Some, for example, float versions of a lineage of Bodhidharma that partially coincide with the version that became fixed as historically accurate in the Song. Others espouse doctrines (e.g., “sudden awakening”) that came to be taken for granted as orthodox in the traditional histories. The authors of the proto-histories, moreover, had diverse aims and did not all belong to any one school or movement within the Buddhist sangha in the Tang. In some cases they represented groups that were in direct competition for lay patronage and official recognition.

The class of modern histories is the most heterogeneous of all. It comprises a wide range of historiographical genres and approaches, such as: the biographical study of individual Chan masters; the philological and philosophical study of uniquely Chan genres of literature; the intellectual history of Chan doctrines and semi-mythical narratives; the ethnographic history of Chan myths and rituals as the product of the adaptation of Indian Buddhism to the Chinese cultural milieu; and the social and religious history of Chan institutions and practices. As the mythological and religious dimensions of the traditional histories have become clearer to modern scholars, moreover, they have struggled to redefine the Chan Lineage/School in ways that are more consistent with scientific methods and critical theories. Those efforts have contributed to the heterogeneity of the field, and have also resulted in widespread (but not widely acknowledged) confusion about its object of study.


The Historiographical Stance of this Book


In one sense, this book — Histories of Chan (Zen) — can be viewed as just one more in a long line of contributions to the modern field of Zen studies. After all, while my system of classifying them is new, the texts that I deal with under the rubrics of “traditional histories” and “proto-histories” have already been identified and used as primary sources by others who have preceded me in that field. I am, of course, deeply indebted to all those predecessors, for most of what I know about the history of Chan Buddhism in China was learned directly from them, or from sources that they collected, edited, and published. Nevertheless, because I call some of the fundamental assumptions, categories, and methods of the modern field of Zen studies into question, I do distance myself from it in important ways.

In the preceding section I describe this book as a “history of histories” because large parts of it — Parts Two, Three, and Four — are dedicated to surveys of what earlier historiographers, both traditional and modern, have written about “Chan” (variously defined). In Part Five, however, I take the documentary materials analyzed in the first four parts of the book and reshape them into a new, more inclusive and coherent historical narrative, using my own stipulative definition of the Chan School as an organizing principle.


Deconstruction vs. “Master Narratives”

Having spent so much effort “deconstructing” both traditional histories of the Chan Lineage and modern histories of the Chan Lineage/School, it might seem strange that I should turn around and present my own positive account of the history of the Chan School, as if I had some more privileged access to what “really happened.” I do, in fact, produce in this book what the deconstructionists who flourished in academia in the 1980s and beyond would disparagingly (as if it were the worst kind of intellectual faux pas) call a “master narrative,” and I make no apologies for it. Indeed, I would argue that it is in the very nature of stories (even those that are ironic, apophatic, purposefully ambiguous, or intentionally deceptive) to strive to be, or at least to look like, the “final word” on the subjects they address.

No narratives, of course, not even the most rigorously tested and pragmatically reliable scientific principles and theories, ever achieve that goal in any permanent way. For one thing, they all are immediately susceptible to rebuttal, revision, refinement, and other forms of criticism. And, it is just a matter of time before they are consumed (in toto or piecemeal) and digested by other self-aggrandizing narratives, or perhaps just left to rot and disintegrate. Despite the ultimate failure of all narratives, it remains a vitally important exercise, not a vain one, to judge whether the stories we encounter are relatively true or false, complete or incomplete, coherent or garbled, elegant or sloppy, relevant or irrelevant, helpful or useless, and so on.

Critical theory in the West has belatedly come to an understanding that has long been expressed in the literature of Mahāyāna Buddhism: in the final analysis even “real” entities are in some sense a product of the imagination, and even “imaginary” entities have a reality akin to that of other mental constructs. Post-modern theorists can still learn a lesson from Mahāyāna philosophy, however, which is that the distinction we habitually draw between real and imaginary entities is an immensely useful and successful conceptual model for navigating the world — an expedient device (fangbian 方便, S. upāya) that is not likely to be improved upon anytime soon. Its corollary in speaking about the past, of course, is the distinction between “myth” and “history.” That also breaks down when subjected to rigorous critical analysis, but despite its weaknesses, it remains indispensable to the historian.

It is true that all narratives, whether we label them as “fact” or “fiction,” are ultimately fictive and inherently biased, but that does not mean that they are all equally effective (or ineffective) as devices for explaining what happened in the past, predicting what will happen in the future, and guiding actions in ways that are more likely to have successful (i.e., desired) outcomes. Thus, I view the naive nineteenth-century positivism embraced by pioneering modern historians of the Chan Lineage/School as an improvement over the theology (or “buddhology”) that informs the traditional histories, but one of my main objections to the modern scholarship is that it has not been rigorous enough in weaning itself from reliance on unscientific categories such as the traditional notion of “lineage.”

I do not maintain that my historical narrative about the rise of the Chan School in medieval China is based on any “facts” or “events” per se that exist independently of people’s telling of them. I only claim that I adhere to current rules of critical historiography somewhat better than story-tellers who have came before me; that I rectify some points of inadvertent ambiguity in their accounts; and that my narrative is limber and coordinated enough to embrace all of the existing documentary evidence without mishandling or dropping any of it. Like the Mahāyāna philosophers who argued that all discourse takes place only on the level of conventional truth, I readily concede that my definition of the Chan School is ultimately arbitrary, and that my account of its history is a biased construct. Nevertheless, until someone comes along to produce a more complete and satisfying narrative, the one I present here should stand as historically “true.”


A New View of the Chan School

Scholars in the modern field of Zen studies have generally regarded the Chan Lineage/School as a single, coherent religious movement — identifiable by some distinguishing doctrines, practices, or social arrangments — that flourished in the Tang and continued to exist, mutatis mutandis, down through the Song and the Yuan. I do not share that view. I see no evidence of any unbroken tradition of distinctively “Chan” doctrines, practices, or institutions that existed in the Tang and continued on into the Song.

What I do see is a narrative continuum, extending from the proto-histories of the Tang down through the traditional histories of the Song and Yuan, in which stories of a lineage founded by Bodhidharma were repeatedly picked up and retold, elaborated on in this way or that, and used for religious legitimation and political gain by different groups of monks within the Buddhist sangha at different times and places. A narrative tradition of this sort differs from a tradition of social organization or material culture in that it does not depend on any living community to sustain it from one generation to the next. Stories can survive in books over long periods of time, even when nobody reads them, to be picked up later and told again, or woven into other historical narratives. Stories can also travel in books and be adopted by far-flung communities that have little or no connection to the original authors. The proto-histories provide evidence that this happened in Tang China with stories of Bodhidharma’s lineage.

There were a number of different schools of Buddhism in the Tang, some more localized than others, that sought legitimacy and lay patronage by retrospectively forging genealogical connections to Bodhidharma. Leading examples include: the school of Faru 法如 (638-689), the Northern School of Shenxiu 神秀 (606?–706), the Oxhead School of Fachi 法持 (635-702), the Southern School of Heze Shenhui 荷澤神會 (684–758), the Jingzhong School of Wuxiang 無相 (694–762), the Baotang School of Wuzhu 無住 (714–774), and the Hongzhou School of Mazu 馬祖 (709-788). Although modern scholars, following the lead of the traditional histories, regard all these schools as branches of a single “early Chan Lineage/School,” the schools did not understand themselves in that way. They tended, rather, to claim that they were the only true heirs of Bodhidharma and to cast others who made similar claims either as illegitimate cousins or outright impostors.

Contrary to what some modern scholars still believe, the historical evidence does not support the notion that any of the aforementioned schools were actually descended from the Indian monk named Bodhidharma (whose biography appears in Daoxuan’s Additional Biographies of Eminent Monks) in the sense of perpetuating identifiable doctrines, practices, or social arrangments that he had originally embraced. I prefer to treat those schools as separate entities because, as best we can tell from the proto-histories, they differed significantly from one another in their teachings and practices and had little in common save the aforementioned strategy of seeking legitimation through genealogy.

I stress the discontinuities between the various schools in the Tang that contributed to the story of Bodhidharma’s lineage because that is needed as a corrective to the traditional view, still accepted in the modern scholarship, that they were all part of a single extended clan whose internal disputes over inheritance were similar to ones that might set siblings, cousins, or second cousins against each other in the succeeding generations of any family whose founding patriarch had amassed great wealth. I hasten to qualify my statements, however, by further stressing that all of the schools named above were comprised of monks who belonged to one and the same state-controlled Buddhist sangha in the Tang. Thus, while they were not related to one another on account of any actual descent from Bodhidharma, they were closely related by their common participation in a monastic institution that imposed a fairly uniform set of moral rules and procedural guidelines for individual and communal living. That institution, which also included many schools of Buddhism that made no use of a Bodhidharma genealogy (e.g., Huayan 華嚴 and Tiantai 天台), did continue to exist, albeit with evolutionary changes, down through the Song and the Yuan.

It was early in the Song that a high degree of centralized control began to be exerted over the various stories of Bodhidharma’s lineage that had been inherited from the Tang and Five Dynasties period. Those were collected, collated, censored, edited for genealogical and stylistic consistency, and published in imperial editions of the Buddhist canon. When all of the diverse lineage claims involving Bodhidharma received from the past were organized in the traditional histories, the vast multi-branched Chan Lineage of ancestral teachers that appeared in those genealogies gripped the imagination of Chinese Buddhists and gained in prestige. Buddhist monks and their lay supporters began to use the genealogical records to forge their own links to Bodhidharma and claim for their teachers and themselves the distinction of being dharma heirs in his lineage, which was said to represent a “singular transmission” (danchuan 單傳) of the awakening of the Buddha, a “transmission of mind by means of mind” (yixin chuanxin 以心傳心) down through the generations, as opposed to a transmission that relied on scriptures. They collectively venerated the ancestral teachers who were memorialized in the traditional histories, and they treated the sayings attributed to those ancestors as a kind of sacred discourse worthy of endless emulation and commentary.

Monks who could convincingly demonstrate that they were heirs to the Chan Lineage, using genealogical records and certificates of dharma inheritance (sishu 嗣書) that they received from their own teachers to make that case, gained privileged access to the abbacies of the leading, state-supported Buddhist monasteries. Abbots who were Chan masters specialized in ritual reenactments of the distinctive modes of rhetoric and gesticulation that were modeled in the traditional histories. Formal rites of dharma transmission, the bestowal of inheritance certificates upon dharma heirs, and the ongoing maintenance of genealogical records became the norm in elite monastic circles. Those practices and the texts that informed them all served to distinguish the Chan School and to bind people together as its members. They were also the basis for the transmission of a clearly identifiable Chan School to the rest of East Asia, where it thrived as Sǒn, Zen, and Thiến.


Go to top


[1] I use the term “medieval China” advisedly herein, as a shorthand way of referring collectively to the Tang, Song, and Yuan dynasties. My usage of “medieval” does not comport with periodization schemes currently in fashion among historians of China, who are not accustomed to grouping those dynasties together.

[2] The Japanese word zen 禪 originated as a borrowing (with a slightly altered pronunciation) of the Chinese chan 禪. There is a great deal of overlap in their respective meanings, but it should not be assumed that everything called Zen in Japan was necessarily regarded as Chan in China. For example, styles of ink painting, calligraphy, and landscape gardening that were cultivated by literati elites in Song China, many who of whom were neither Buddhists nor sympathetic to Chan, later became known in Japan as “Zen” arts.

[3] Sōtō is the Japanese pronunciation of the Chinese Caodong 曹洞; Rinzai is the Japanese pronunciation of the Chinese Linji 臨濟.

[4] Bunkachō 文化庁, ed., Shūkyō nenkan, 2003 Edition 宗教年鑑, 平成15年版, 46-47, 70-71.

[5] Bunkachō 文化庁, ed., Shūkyō nenkan, 2003 Edition 宗教年鑑, 平成15年版, 70-71.

[6] Chogye is the Korean pronunciation of the Chinese place name Caoxi 曹溪; it refers both to the sixth ancestor of the Chan lineage, Huineng, who was known as the “great master of Caoxi,” and to an influential monastery founded by the Korean monk Chinul.

[7] Sǒn 禪 is the Korean pronunciation of the Chinese word chan 禪. It covers much of the same semantic range as its Chinese counterpart, but due to historical differences in usage it should not be considered a perfect synonym.

[8] For a detailed account of Sǒn Buddhism in contemporary Korea, see Robert E. Buswell, Jr., The Zen Monastic Experience (Princeton: Princeton University Press, 1992).

[9] Chinul was especially influenced by the writings of Guifeng Zongmi 圭峰宗密 (780-841), Yongming Yanshou 永明延壽 (904-975), and Dahui Zonggao 大慧宗杲 (1089-1163). He gave the name Mt. Chogye (Chogyesan曹溪山) to the Songgwang Monastery (Songgwangsa 松廣寺) that he founded.

[10] Thiến 禪 is the Vietnamese pronunciation of the Chinese word chan 禪.

[11] Lâm-te is the Vietnamese pronunciation of the Chinese name Linji 臨濟.

[12] For a history of the movement, see Jiang Wu, Enlightenment in Dispute: The Reinvention of Chan Buddhism in Seventeenth-Century China (New York: Oxford University Press, 2008).

[13] Holmes Welch, The Practice of Chinese Buddhism: 1900-1950 (Cambridge: Harvard University Press, 1967), 281.

[14] Because the modern study of the history of the Chan Lineage/School began in late nineteenth and early twentieth Japan, where it was known as “Zen studies” (Zengaku 禪學), and because Japanese scholarship has continued to dominate and define the modern field down to the preent day, I refer to that field as “Zen studies” (using the Japanese pronunciation of 禪), regardless of whether the object of study is located in China, Korea, Vietnam, or Japan.

[15] The basic meaning of “dharma” (fa 法) in this context is “teaching,” but in the Chan tradition it refers to an insight or understanding that cannot be conveyed in words and has no “signs” (xiang 相) that might be used, in an objective fashion, to ascertain its presence or absence.

[16] T.2008.48. The full title of the text as it was known in the Song is: Dharma Treasure Platform Sūtra of the Sixth Ancestor (Liuzu dashi fabao tanjing 六祖大師法寶壇經).

[17] T. Griffith Foulk, ‘‘The ‘Ch’an School’ and its Place in the Buddhist Monastic Tradition’’ (Ph.D. Dissertation, University of Michigan, 1987), pp. 212-241.

[18] Richard Robinson, Definition (Oxford: Clarendon Press, 1954).

[19] At present, standard introductions to the discipline of logic list “lexical” and “stipulative” as the first two of five basic kinds of definition, the other three being “precising,” “theoretical,” and “persuasive” definitions. See the Brittanica “Philosophy Pages,” s.v. “Logic,” s.v. “Definition and Meaning,” @ (website active 3/18/2011). I do not know if Robinson coined the terms “lexical definition” and “stipulative definition,” but I learned them from his book entitled Definition.

[20] Robinson, Definition, p. 8.

[21] Robinson, p. 16.

[22] Ibid, p. 154.

[23] Ibid, p. 155.

[24] Robinson’s position is that the notion of real definition is a confusion of at least twelve distinct activities, some fundamentally valid and some invalid, and that “we had better drop the term ‘real definition,’ and call each of the twelve different activities that ‘real definition’ has meant by a more specific name….” The twelve activities he identifies are: 1) Searching for an identical meaning in all the applications of an ambiguous word; 2) Searching for essences; 3) Describing a form and giving it a name; 4) Defining a word, while mistakenly thinking that one is not talking about words; 5) Apprehending a tautology determined by a nominal definition; 6) Searching for a cause; 7) Searching for a key that will explain a mass of facts; 8) Adopting and recommending ideals; 9) Abstracting, i.e. coming to realize a form; 10) Analyzing, i.e. coming to realize that a certain form is a certain complex of forms; 11) Synthesizing, i.e. coming to realize that a certain form is a certain part of a certain complex form; 12) Improving one’s concepts. (Ibid, pp. 189–190).

[25] The explanatory examples I present herein are all my own, not Robinson’s.

[26] Other examples of null sets are geometrical figures, such as circles or spheres, as those are defined in mathematics. Things in the real world can be judged more or less circular or spherical, but none actually satisify the definitions perfectly.

[27] Sekiguchi Shindai, “Zenshū no hassei,” in Fukui Hakushi Juju Kinen Ronbunshū Kankōkai, ed., Fukui hakushi juju kinen tōyō shisō ronshū (Tokyo: 1960), 321-338.

[28] Yanagida Seizan, Shoki zenshū shisho no kenkyū (Kyoto: Hōzōkan, 1967), 437-460.

[29] This distinction between “lineage” and “school” is one that I stipulate for the purposes of this book and recommend for future use in the field of Zen studies. I am fully aware that there is nothing in the ordinary, lexically defined English meanings of the words “lineage” and “school” that necessarily associates the former with imaginary social entities or the latter with real ones.

[30] Yanagida Seizan, ed., Sodō shū, Zengaku sōsho No. 4 (Kyoto: Chūbun, 1984). Also, CBETA.B25.144.

[31] T.2076.51.196b-467a. Henceforth referred to as the Jingde Record.

[32] Thus, for example, Kinsei zenrin sōbōden 近世禪林僧寶傳 was recently reedited and published by the Zen Bunka Kenkyūsho (The Institute for Zen Studies) at Hanazono University in Kyoto. An electronic edition of the same text was also made available in the ZenBase CD1 (1985), edited and produced by Urs App at the International Research Institute for Zen Buddhism at Hanazono College.

[33] Bernard Faure, Chan Insights and Oversights: An Epistemological Critique of the Chan Tradition (Princeton: Princeton University Press, 1993), 101-104.

[34] Suzuki Teitarō 鈴木貞太郎, who was given the dharma name (hōgō 法號) of Daisetsu 大拙 or “Great Bungler” by his Zen master, Shaku Sōen 釋宗演 (1859-1919).

[35] The expression “real world,” of course, is just another linguistically formed category that unavoidably entails reification and gross over-simplification. In the Mahāyāna view, nothing ultimately true can be said of the “real world.” The closest one can come to a true statement is the tautology, “It is the way it is” (rushi 如是, S. tathatā).

[36] The Platform Sūtra survives in at least eleven different recensions, all of which have been published in Yanagida Seizan, ed., Rokuso dankyō shohon shūsei, Zengaku sōsho 7 (Kyoto: Chūbun, 1976). The oldest is the Dunhuang manuscript edition, which dates from sometime between 781 and 801; for these dates, see John R. McRae, “Yanagida Seizan’s Landmark Work on Chinese Ch’an,” in Cahiers dExtrême–Asie (Bilingual Journal of the Ecole Francaise d’Extrême–Orient) 7 (1993–1994): 75; also Philip B. Yampolsky, The Platform Sūtra of the Sixth Patriarch, 90-98). It was discovered in the Stein collection at the British Museum by Yabuki Keiki in the mid-1920s and published in 1928 in the Taishō edition of the Buddhist canon (T.2007.48:337a-345b); for an English translation, see Yampolsky, The Platform Sūtra. The rescension of the text that circulated widely in the early Song was one edited by Huixin 恵昕 in 967, which is now lost; the extant text that is probably closest in contents to it is the Kōshōji edition, discovered by Suzuki Daisetsu at the Kyoto monastery by that name in 1934.

[37] Henceforth referred to as the Baolin Biographies. Yanagida Seizan, ed., Sōzō ichin: Hōrinden, Dentō gyokuei shū, Zengaku sōsho 5 (Kyoto: Chūbun, 1983). Tanaka Ryōshō 田中良昭, Hōrinden yakuchū 宝林伝訳注 (Tokyo:Nakayama Shoten, 2003). A digital version of the text is found in the CBETA Supplement to the Tripiṭaka (大藏經補編), B 81:4, under the full title of Shuangfengshan caohouxi baolin zhuan  雙峰山曹侯溪寶林傳, which is based on the text published in the Complete Works of the Chan School, Canzong quanshu 禪宗全書, under the general editorship of Lan Jifu藍吉富, in the first volume of the histories and biographies section (史傳部) (Taibei: Wenshu chubansha, 1988).

[38] ZZ.110.

[39] T.2015.48.

[40] For a concise bibliography of texts by Shenhui, see Philip B. Yampolsky, The Platform Sūtra of the Sixth Patriarch (New York: Columbia University Press, 1967), 24-25, note 67; also Yokoi (Yanagida) Seizan, “Tōshi no keifu,” in Nihon bukkyō gakkai nenpo 19 (April, 1954).

[41] T.2838.85:1291a-c. For a critical edition and annotated Japanese translation, see Yanagida Seizan, ed., Shoki no zenshi I: Ryōga shiji ki, Denbōhōki, Zen no goroku 2 (Tokyo: Chikuma, 1971), 330-435. For an English translation see John McRae, The Northern School and the Formation of Early Chan Buddhism (Honolulu: University of Hawaii Press, 1986), 255-269.

[42] T.2837.85:1283a-1290c. For a critical edition and annotated Japanese translation, see Yanagida Seizan, ed., Shoki no zenshi I: Ryōga shijiki, Denbōhōki, Zen no goroku 2 (Tokyo: Chikuma, 1971), 48-326. An English translation, entitled “Records of the Teachers and Students of the Lanka,” may be found in J.C. Cleary, Zen Dawn: Early Zen Texts from Tun Huang (Boston & London: Shambala, 1986), 19-78.

[43] T.2075.51:179a-196b; for a critical edition and annotated Japanese translation, see Yanagida Seizan, ed., Shoki no zenshi II: Rekidai hōbōki, Zen no goroku 3 (Tokyo: Chikuma, 1971), 39-324. For a condensed English translation of the section on Wu-chu, see Jeffrey Broughton, “Early Chan Schools in Tibet,” in Robert M. Gimello and Peter N. Gregory, eds., Studies in Chan and Hua-yen (Honolulu: University of Hawaii Press, 1983), 19-29.

[44] First published in Suzuki Daisetsu, ed., Tonkō shutsudo shōshitsu issho 敦煌出土少室逸書 (Kyoto: privately published collotype photograph edition, 1935). There are now nine Dunhuang manuscripts that contain parts of this text; for a list of those see Jeffrey L. Broughton, The Bodhidharma Anthology: The Earliest Records of Zen (Berkeley: University of California Press, 1999), 121, note 12. For an annotated edition based on six of those manuscripts and one found in a 1907 Korean collection, see Yanagida Seizan, ed., Daruma no goroku: Ninyū shigyō ron, Zen no goroku 1 (Tokyo: Chikuma, 1969). For an English translation based on that edition, see. Broughton, The Bodhidharma Anthology, 8-52.

[45] ZZ 2-15-5.405a-408b; T #2009.48.373b12-376b14.

[46] ZZ 2–15–5.408c–411b; T #2009.48.370c11-373b11.

[47] ZZ 2-15-5.411c-414c; T #2009.48.366c18-369c18.

[48] T.2009.48:370a29-370c10.


Work in progress, not for citation.
No part of this content can be used, distributed, and reproduced without written consent from the author.


Go to top