Dharma · AI Companion

祖師

「祖师」指禅宗传承历代法脉之宗师,亦可指当下问答中可被勘验指认之前人。AI-generated

Type PhenomenaDifficulty IntroductoryTruth-level span outside the four-fold two-truths frameworkInitial AI estimate · evolves with use

A Contemporary ReadingAI-generated

1. What this concept is actually saying

The first thing to notice when approaching 祖師 through the three source texts — 《古尊宿語錄》 (Records of Ancient Worthy Monks), 《景德傳燈錄》 (Jingde Records of the Transmission of the Lamp), and 《明覺禪師語錄》 (Recorded Sayings of Chan Master Mingjue) — is that the sources never define 祖師. They use it. Across the three, the term is never unpacked as a doctrinal category. Instead, it is set in motion inside encounters: a monk returns from bowing at a stupa, and the teacher asks, "祖師道什麼" — "what did the patriarch say?"; a monk says there is a 祖師 in the hall, and the teacher replies, "喚來與老僧洗脚" — "call him over to wash the old monk's feet." Definition yields to test. Naming yields to confrontation. To get at what 祖師 actually is, we have to follow the threads the sources leave us.

Thread one — 祖師 as the lineage-positioned transmitter of the dharma. This is the most familiar layer, and 《景德傳燈錄》 develops it most fully. The text counts: "自迦葉至般若多羅,有二十七祖" — "from Kāśyapa to Prajñātāra, twenty-seven patriarchs"; with side-branches from Bodhidharma's line, "二十二人,總有四十九祖" — "twenty-two, totaling forty-nine patriarchs"; not counting branches, "凡三十七世" — "thirty-seven generations" from the Seven Buddhas to the Fourth Patriarch. 祖圖 (ancestral diagrams) already circulated, recording "五十餘祖" — "more than fifty names" — but the compiler himself admits that "支派差殊宗族不定,或但有空名者" — "branches differ, lineage unsettled, some are only empty names." Even the official genealogy is provisional, contested, half-mythic. The text then tells the story of 曇曜 (Tanyao) during the Northern Wei persecution: he recorded the patriarchs' names on silk, sewed the list into his collar, hid in a cave for thirty-five years, and only emerged under Emperor Wencheng to "集諸沙門重議結集" — "gather the monastics to re-discuss and re-compile." The transmission of the lineage itself had to be smuggled through catastrophe. So at this most "official" layer, 祖師 is not a stable list of saints; it is a record repeatedly salvaged and openly acknowledged as still unsettled.

Thread two — 祖師 as the living present-tense test. This is the layer that 《古尊宿語錄》 insists on. Here 祖師 is not someone who died a thousand years ago; 祖師 is whoever can be pointed at in the meditation hall *right now*. A monk comes back from 南華 (Nanhua, where the Sixth Patriarch's stupa stands), and the teacher doesn't say "ah, the Sixth Patriarch." He says, "祖師有什麼言句" — "what words did the patriarch say?" and warns "不得錯舉" — "don't misquote." A monk says "禮塔來" — "I came from bowing at the stupa," and the teacher asks, "祖師道什麼" — "what did the patriarch say?" The implicit message: your pilgrimage means nothing if you cannot produce the actual pointing. Then comes the deeper move: "還見祖師麼" — "did you see the patriarch?" The monk responds, the teacher says "用見作什麼" — "what's the use of seeing?" and "既去無過,見有什麼過" — "you went, no fault; what's the fault in seeing?" The "seeing" of the patriarch has been turned into a present-tense exchange. And most cuttingly: "堂中還有祖師也無" — "is there a 祖師 in the hall?" A monk answers "有" — "yes." The teacher says "喚來與老僧洗脚" — "call him over to wash the old monk's feet." The question of whether a real 祖師 is present in the room right now is not metaphysical; it is practical — can anyone here actually transmit, or only claim to? The 祖師 is exposed as a function, not a statue.

Thread three — 祖師 as transmitter of heart-dharma (心法), not textual dharma. 《景德傳燈錄》 records the moment of transmission itself. The teacher, about to hand on the lineage, says: "吾若言有法與人,即為誑汝。但且隨方解縛假名三昧" — "If I said I had a dharma to give to people, I would be deceiving you. Just untie bonds according to circumstance; provisionally it's called samādhi." What is transmitted is not a doctrine, not a method, not a scripture; it is the capacity to release another being's bondage, in whatever form the situation calls for. The famous exchange on "祖師西來意" (the patriarch's meaning in coming from the West) — "如何是祖師西來意。師曰。有力者負之而趨" — "What is the patriarch's meaning in coming from the West? The master said: the strong one carries it and runs" — captures the same point: what came west was not a teaching to be passed down intact; it was a force carried by those with the strength to carry it. The carrier is the 祖師; the carrying is the lineage.

Thread four — 祖意 (the patriarch's intent / pointing), a related but distinct term. The verse from 《古尊宿語錄》 reads: "昔人曾指出,今古道傳馨。祖意憑君悟" — "People of old once pointed it out; through ancient and modern, the way transmits its fragrance. The patriarch's intent, you must awaken to yourself." 祖意 is the pointing that previous 祖師 have already made, the resonance still carrying through the tradition — but the actual awakening to it cannot be handed over; it must be "憑君悟" — "entrusted to your own awakening." This is why 祖意 is not a doctrine you can study into yourself. You can study all the 祖師's words and remain untouched by 祖意.

Thread five — traces and words (遺蹟, 言句): what is left behind and what cannot be grasped. The famous line "祖師遺下一隻履,直至如今覓不得" — "The patriarch left behind a single sandal; until now, no one can find it" — uses the image of a sandal. A sandal is a trace of someone who passed through: the most humble, intimate object, the thing that touches the ground between your foot and the world. The 祖師's trace is real; the tradition remembers it; but you cannot track the 祖師 down through the trace the way you can track a person through a footprint. The trace is a pointer, not a map. The same text records the monk who comes from the lineage hall at Nanhua and is asked, "祖師有什麼言句" — "what words did the patriarch say?" with the warning "不得錯舉" — "don't misquote." The patriarch's words exist; they can be transmitted, and they are tested for accuracy. But then comes the disarming: "祖師挂什麼衣" — "what robe does the patriarch wear?" The answer: "終日著衣未曾挂著一縷絲" — "wore clothes all day, never hung on a single thread." The robe, the form, the visible marker of being a 祖師 — none of that is what makes the 祖師. The 祖師 wears, but is not hung up on, the form.

Thread six — where the 祖師 doesn't reach. The single most striking formulation in 《明覺禪師語錄》 is: "祖師不到處,時人知有;時人不知處,過在祖師。作麼生辯。不得春風華不開" — "Where the patriarch does not reach, ordinary people know there is; where ordinary people do not know, the fault lies with the patriarch. How to discern? Without the spring wind, flowers do not open." This is a structural admission, not a scandal. 祖師 are real transmitters, but they do not cover everything. There are regions they don't reach — and ordinary people, in their very ordinariness, can know that. There are regions ordinary people don't know, and that gap is on the 祖師, not the layperson. The "作麼生辯" — "how to discern?" — refuses a doctrinal answer. The 代語 (substitute answer) given is: "不得春風華不開" — "without the spring wind, the flowers do not open." The spring wind is the 祖師, the flower is the student, and the opening is the student's own actual response. You don't "know" the 祖師's intent by knowing about them; you open, or you don't. This is also why the same text says "箇箇道我會" — "everyone says 'I understand'" — and then "會即且致,作麼生舉" — "set aside the 'understanding' for a moment; how do you lift it up / present it?" The 祖師 is the one who can be tested by being asked to actually present, not the one who can recite.

Putting the threads together. 祖師 is therefore not a noun-thing but a multi-layered function: (a) the lineage-positioned transmitter, (b) the present-tense test, (c) the carrier of an un-graspable pointing, (d) the person whose traces and words can be cited but not converted into a system, (e) the one whose intent (祖意) lives in actual awakening, not in records, and (f) someone who, by the tradition's own admission, does not and cannot cover everything. Each of these threads is not a definition; it is a direction the tradition pulls in. The sources are careful not to let the reader settle on any single one.

Distinctions the sources insist on (and that the reader must not collapse). - 祖師 is not 宗師. 《明覺禪師語錄》's preface says "有愧於宗師" — "I am ashamed before my own teacher" — and 宗師 here is the personal honorific for one's living teacher. 祖師 is lineage-positioned; 宗師 is the one you actually face and learn from. The preface-writer is not a 祖師; he is a student of one. - 祖師 is not 禪祖 / 祖 in a new way. 《景德傳燈錄》 uses 禪祖 and 祖 as shorter variants: "我師禪祖中當得為第八" — "my teacher should rank as the eighth among Chan patriarchs." Stylistic variants, not new concepts. - 祖師 is the source; 法嗣 (dharma heirs) are the receivers. The 《景德傳燈錄》 organizes itself by 法嗣: "南陽慧忠國師法嗣五人" ("National Teacher Huizhong of Nanyang, five dharma heirs"), "洛陽荷澤神會大師法嗣一十八人" ("Master Shenxiu of Heze at Luoyang, eighteen dharma heirs"). Every 祖師 is a node radiating a network of heirs; every heir is potentially a future 祖師. The lineage is not a closed list of saints but a relational fabric. - 祖師 is not 祖庭 (ancestral temple). 祖庭 is the place; 祖師 is the person-and-function. The sources repeatedly test those who come from 祖庭, but they test them as if the place means nothing without the function being present and verifiable. - 祖師 is not 祖意. 祖師 is the transmitter; 祖意 is the pointing, which the transmitter carried and which you must awaken to yourself.

These distinctions are not pedantic. They are the actual scaffolding the sources build on. Treat 祖師 and 宗師 as the same, and the meditation-hall exchange "堂中還有祖師也無" / "喚來與老僧洗脚" collapses. Treat 祖師 and 祖庭 as the same, and the warning "禮塔來" / "祖師道什麼" makes no sense. Treat 祖師 as a person and 祖意 as a doctrine they taught, and "祖意憑君悟" loses its edge.


2. A walk-through in contemporary life

The point of this section is to map the threads above — strictly, without smearing — onto scenes a contemporary reader will recognize. I will walk thread by thread, and at the end call out where the popular mapping goes wrong.

Scene A — the "ancestral diagram" of a software project. Open any large open-source project — a major programming language, a foundational library. The README lists the original authors. The git log shows a chain. There are roughly twenty-to-thirty "founding" names the community can recite, with side-contributors that may or may not be canonical. Conferences have "hall of fame" slides. The 《景德傳燈錄》's 27/49/37 counting is exactly this: a list, with branches, with the compiler's own admission that "支派差殊宗族不定,或但有空名者" — "branches differ, the genealogy is unsettled, some are only empty names." The names in the README are also not fully settled; some are inflated, some are forgotten, the count shifts. And the story of 曇曜 hiding the names on silk in his collar for thirty-five years is the story of every project whose copyright notices, license files, and AUTHORS.md have to be smuggled through hostile environments — layoffs, acquisitions, hostile forks, regime change. The first thread of 祖師 is here: lineage-positioned, with a count that is partly real and partly provisional.

The walk-through check: if you treat the README names as the 祖師, you stop at hagiography. The 祖師-test in the source texts is the moment when a developer who can quote the README cannot actually debug the system. The README is the "一隻履" — "a single sandal" — the trace that proves the founder passed through but does not let you find them. The 祖師 as lineage is real; the 祖師 as a name in a file is a fossil.

Scene B — the job candidate who lists "mentored by." A candidate's CV names a famous mentor, a reference, a "祖師" they can cite. The interviewer — acting here like the Chan teacher — is supposed to ask not "who was your mentor?" but "what did your mentor say, exactly?" — i.e., "祖師有什麼言句" with the warning "不得錯舉." If the candidate can name the mentor but cannot produce a single accurate pointing, the mentor was not a 祖師 for them, only a credential. The 祖師 as present-tense test is the moment when lineage becomes lineage, or stops being lineage.

The walk-through check: this is where most contemporary readers map the concept wrongly. They identify 祖師 with the famous name. The source texts do the opposite: the famous name is necessary as orientation but is not where the test happens. The test happens in the second question, the one that demands you produce what was actually transmitted. (The popular Bodhidharma or Huineng or Zhaozhou as objects of devotional consumption in modern Buddhism is 祖師 reduced to a name. The sources would treat that as a fossil, not a transmission.)

Scene C — the meditation app and the dharma teacher. A contemporary reader sits in front of a screen — a meditation app, an online dharma talk, a YouTube channel with a teacher who has a large following. The teacher may be wise, may be teaching exactly what the tradition teaches, may even be a credentialed lineage holder. The Chan source-text test: "堂中還有祖師也無" — "is there a 祖師 in this hall right now?" If yes, the next move is "喚來與老僧洗脚" — "call them to do the next ordinary thing." In the app, the equivalent test is: when you close the app, when you go wash the dishes, when you have a fight with your partner — does whatever was transmitted actually show up? If the dharma stays on the cushion, you had a teacher; you did not have a 祖師. The 祖師 as a present-tense test is not about how wise the teacher appears; it is about whether the transmission survives ordinary life.

The walk-through check: the 祖師 as function is exactly this — a transmitter whose effect is verifiable outside the room of transmission. A teacher who is profound in the talk and unrecognizable in the supermarket has not yet been tested as a 祖師 by this tradition. (This is not a guarantee; the sources are clear that testing can fail in both directions. But the test is the test.)

Scene D — the "where the patriarch doesn't reach" — the AI question. A contemporary reader is asking an AI, or being asked, about the meaning of practice, the transmission of dharma, the lineage of teachers. The AI can recite the patriarchs; it can produce a perfect 祖意, can summarize 不得春風華不開, can output "有力者負之而趨" with proper attribution. The Chan sources' diagnosis of exactly this situation is encoded in "祖師不到處,時人知有" — "where the 祖師 doesn't reach, ordinary people know there is." A region of experience exists where the historical 祖師, the lineage, the words, the diagrams simply do not reach. The contemporary reader can know this region exists — that the AI can simulate a 祖師 without being one, that the lineage cannot pre-cover every new configuration of suffering and question. The sources do not lament this; they name it. The "過在祖師" — "the fault lies with the 祖師" — half of the same sentence is also crucial: in the regions where ordinary people don't know, the responsibility for not-knowing falls on the 祖師, not on the layperson. The 祖師 is responsible for transmitting; where they have not, the gap is theirs. The contemporary analog: the lineage's responsibility is to actually engage, not to declare that the new questions are "outside the dharma."

A note of caution about the AI comparison: this is a parallel, not an equivalence. The sources do not "predict" AI; they describe a structural feature of transmission — that any lineage will leave gaps — and that feature maps onto the present AI moment usefully, but the sources have nothing to say about machine learning as such. It would be a category error to claim "Chan anticipated AI." What Chan does is offer a discipline for thinking about gaps in transmission; that discipline happens to apply.

Scene E — the "千徒不共知" — the thousand disciples who do not all know. A modern teacher, a famous spiritual author, a popular podcaster — they have thousands of followers. The 明覺禪師 verse "人誰訪祖師" / "千徒不共知" — "who visits the 祖師? A thousand disciples do not all know" — is not cynicism. It is structural. Of the thousand who follow, only some are actually visiting. Most are visiting a name, an image, a "thing I have heard about the 祖師." "千徒不共知" doesn't mean most are insincere; it means the transmission is not a broadcast. Each person has to do the visiting themselves. "早晚歸林下" — "sooner or later, return to the forest" — is the suggestion that the time will come when you stop following the broadcast and do the actual visit.

Scene F — the single sandal. A friend dies. You inherit a notebook, a piece of clothing, a sentence of theirs that has stayed with you. The notebook is "祖師遺下一隻履" — the sandal the 祖師 left behind. It is real. It is theirs. You cannot find them through it. Years later, you reach for the notebook in a hard moment, and what helps is not the words but the way the words point — the 祖意 carried in the trace. The 祖師 as trace is a real human experience: someone whose presence lingers in a way that points but does not substitute. The contemporary mistake is to make the sandal into a museum piece (a biography, a brand, a saint). The 祖師-as-sandal is meant to be worn-into-ground, then left.

Scene G — the "我會" trap. A person comes back from a retreat, a course, a year of practice. They say, "I get it." Friends nod; teacher nods. The 明覺禪師 line "箇箇道我會" / "會即且致,作麼生舉" — "everyone says 'I get it'; set that aside, how do you lift it up?" — is the test. The "我會" is fine as a feeling; the tradition tests it by asking you to show the actual move. In contemporary life, this is the difference between saying "I understand boundaries" and being able to actually set one in a tense moment; between saying "I understand impermanence" and being able to actually let go of a relationship that is done; between saying "I understand my patterns" and being able to interrupt one mid-act. The 祖師-test, when it works, is not a quiz. It is a door.

Where popular mappings go wrong (to call out explicitly): - Mapping 祖師 to "famous historical figure" alone is the sandal-as-museum-piece error. The sources never let 祖師 be only historical. - Mapping 祖師 to "my guru" is the 宗師 error. 祖師 is lineage-positioned; the person you face daily is 宗師. Conflating them collapses the test. - Mapping 祖師 to "any skilled teacher" loses the lineage specificity. The 祖師 is a node in a network of pointings, not an isolated sage. - Mapping 祖師 to "I can recite the lineage" is the 我會 error. The lineage can be recited without being carried; carrying is the test. - Mapping 祖師 to "infallible saint" contradicts "祖師不到處" and "過在祖師" directly. The sources are clear: 祖師 have gaps, and where they don't reach, ordinary people can know it.


3. Why a contemporary person needs this concept

The first reason is that the contemporary information environment has *abundance of names, scarcity of transmission*. A modern reader can name Bodhidharma, Huineng, Huangbo, Linji, Dahui — or in other traditions, Aristotle, Aquinas, Kant, Heidegger — and have no access to what any of them actually transmitted. The concept of 祖師 is precisely the discipline of refusing to confuse the name with the transmission. It insists on a test. In a culture that hands out micro-credentials, follow counts, and citation indices as if they were evidence of transmission, this is a needed corrective. The 祖師-test is not a credential; it is a question: "what did the patriarch actually say, and can you produce it, and does it still work?"

The second reason is that the contemporary lineage landscape is structurally different from the Chan landscape, and we need a concept that respects both. We don't have a continuous, named 祖師 → 法嗣 chain in most contemporary domains. We have broken lineages, recovered lineages, appropriated lineages, lineages in the form of citations, lineages in the form of institutional memory. The Chan sources are unflinching about this: even within their own tradition, the lineage was almost lost (曇曜), the count was unsettled, some names were "空名" — empty. The concept of 祖師 is not weakened by lineage-instability; it is built on the recognition that lineage is always at risk. The 祖師, properly understood, is the function that gets tested and re-tested in each generation, not a list that guarantees itself.

The third reason is the asymmetry between what content can be transmitted and what cannot. The Chan sources, especially the line "吾若言有法與人,即為誑汝" — "if I said I had a dharma to give people, I would be deceiving you" — face this head-on. What gets passed on in a tradition is not the dharma itself; it is the capacity to untie bonds, in a form appropriate to circumstance. AI can now transmit content; it cannot transmit capacity. A modern reader surrounded by AI-mediated information needs a way of distinguishing "I have the words" from "I have the capacity." The 祖師-test is one of the few concepts in any tradition that builds this distinction into its very definition: a 祖師 is not the one who has the words; it is the one who can be tested for whether the words have actually opened anything.

The fourth reason is that "祖師不到處" — "where the 祖師 doesn't reach" — is an unfashionable but indispensable concept. Contemporary culture tends to imagine that any gap in a tradition's coverage is a personal failure of the practitioner, or a sign that the tradition is obsolete. The Chan tradition says: 祖師 are real, they cover a lot, and they don't cover everything. Where they don't reach, ordinary people can know it. The "過在祖師" half of the same sentence is even more striking: where ordinary people don't know, the responsibility falls on the 祖師. This is a tradition that takes its own failures seriously, and a contemporary reader who is used to traditions that claim to cover everything (or to disclaim everything as outside their scope) can learn something from a tradition that does neither. The 祖師 is neither omniscient nor irrelevant.

The fifth reason, more quietly, is that the 祖師-concept gives a contemporary person a way of relating to their own teachers, predecessors, and influences that is neither worshipful nor cynical. "祖意憑君悟" — "the patriarch's intent, you must awaken to yourself" — refuses both deference and dismissal. The predecessors pointed. The pointing is real. The awakening is yours. A culture saturated with hagiography on one side and "no one is a real master" cynicism on the other needs the third option the Chan tradition here holds open: lineage is real, transmission is real, and the burden of actual awakening is on you, not on the lineage.

A note of humility: neuroscience has begun to map how skills and dispositions are transmitted between people — mirror neurons, emotional contagion, the way a teacher's regulation of attention can shape a student's over years of close contact. Some of these findings rhyme with what Chan calls transmission, especially the part the sources call 心法 (heart-dharma). But the parallel is partial. Neuroscience describes mechanisms; Chan describes a function-within-a-lineage whose test is whether the student's life actually opens. It would overreach to say neuroscience "proves" Chan, and the sources themselves make no such claim. The honest comparison is: the sciences offer partial accounts of how 祖意 might move between people, and leave the question of whether it actually opens anything still open — exactly the question the Chan sources insist on asking.


4. Common misreadings and clarifications

Misreading 1: 祖師 means "famous old Chinese master." In popular usage, 祖師 has been narrowed to a small handful of mythic figures — usually Bodhidharma and Huineng, sometimes Zhaozhou, Dongshan, or Huangbo. The sources flatly contradict this. In 《景德傳燈錄》, every禅师 is a node in the lineage; the text is organized by 法嗣 networks. The 祖師 is not a closed Hall of Fame; it is the function any lineage-transmitter can occupy. The right reading: 祖師 is a function, not a list.

Misreading 2: 祖師 is the originator or founder of a tradition. This is a particularly sticky error because it makes 祖師 a "first" rather than a node. The Chan sources are clear that the lineage goes back to 迦葉 and ultimately to the seven Buddhas; nobody is the first. Even Bodhidharma is the twenty-eighth or twenty-ninth in the line, depending on how you count. 祖師 is a position in a chain, not a person who started it. The right reading: 祖師 is one who transmits within a chain that did not begin with them.

Misreading 3: 祖師 is the same as 宗師 (one's personal teacher). 《明覺禪師語錄》 itself flags this. The preface uses 宗師 as the personal honorific; the body uses 祖師 as the lineage-positioned transmitter. Conflating them turns a test-of-lineage into a test-of-personal-loyalty, which is the opposite of what the sources do. The right reading: 宗師 is who you face; 祖師 is whose line you are tested in.

Misreading 4: 祖師 is an infallible saint. "祖師不到處,時人知有;時人不知處,過在祖師" is a clear statement that 祖師 have limits and bear responsibility for the gaps in their transmission. The 祖師 is not beyond critique; the 祖師 is the one whose transmission is the most carefully scrutinized. The right reading: 祖師 is a real transmitter, with real gaps, whose gaps are the tradition's problem to face.

Misreading 5: 祖師 is the doctrine the master taught. "祖意" — the patriarch's intent — is the closest thing to a doctrine, and the sources are explicit that it cannot be transmitted as a doctrine: "祖意憑君悟." The right reading: a 祖師 is not the message; a 祖師 is the one through whom the message must be awakened by the student. The 祖師 as a "set of teachings" is a fossil; the 祖師 as a function is alive.

Misreading 6: 祖師 is something only the past has. This is the natural modern error, given that most of the named 祖師 are historical. The 《古尊宿語錄》 material is built precisely to refuse this. "堂中還有祖師也無" is asked about the room that the questioner is currently in. The tradition's claim is that the function is not extinct, only rarer and harder to identify because it is being tested rather than venerated. The right reading: the 祖師 is a present-tense question, not a museum exhibit.

Misreading 7: 祖師 can be reduced to a "guru" or "spiritual master" in the contemporary pop sense. The 祖師 concept is built on a network of 法嗣, a tested lineage, an acknowledgment of limits, and a refusal to substitute form for transmission. The pop "guru" concept typically has none of these features: no tested transmission, no acknowledged limit, heavy reliance on form (robes, titles, charisma). The right reading: 祖師 is the opposite of a guru, in the sense that 祖師 is precisely the role that cannot coast on its own form.

Misreading 8: 祖師 is the one who knows the answers. A persistent reading of Chan is that 祖師 are masters of cryptic answers. The 喚來與老僧洗脚 exchange shows the opposite: when a monk claims to know, the test is whether the knowing survives the next ordinary thing. The 祖師 is not the one who knows the answers; the 祖師 is the one whose transmission can survive the next ordinary question. The right reading: the test is not the brilliance of the answer but the survival of the transmission outside the room of transmission.

Misreading 9: the "I get it" trap. "箇箇道我會" / "會即且致,作麼生舉" — "everyone says 'I get it'; set that aside; how do you lift it up?" This is the most dangerous misreading because it is the easiest to fall into. The contemporary reader can leave a dharma talk, a retreat, a therapy session, a moving book, and feel the "I get it." The 祖師-tradition's response is not to deny the feeling; it is to ask, gently, whether the lift can actually be made. The right reading: the question after every transmission is not "did I get it?" but "what can I actually lift, right now, in my next ordinary moment?" If you can lift nothing, the "I get it" was a feeling, not a transmission.

Misreading 10: 祖師 is the "ancient sage" who is, by virtue of being ancient, more trustworthy than contemporary teachers. The sources do not rank by age. They rank by transmission, which is testable. A young contemporary who can actually untie bonds in the situation in front of them is, in this framework, transmitting the dharma. An ancient name who can be recited but not carried is a sandal, not a transmitter. The right reading: the 祖師 is not a function of age, geography, ethnicity, or era. It is a function of whether the transmission can be tested and survives the test.


A few closing words. The Chan tradition's treatment of 祖師 is, when you sit with it, not hagiography and not critique. It is a sustained discipline of testing. It tests the lineage (does it transmit, or only name?), it tests the contemporary practitioner (can you be the one called to wash the old monk's feet?), it tests the past 祖師 themselves (where did they not reach, and what do you, ordinary person, know about those gaps?), and it tests the student (what can you actually lift, beyond saying "I get it"?). The reason the concept is worth reading carefully today is that very few traditions have built the test into the term itself. Most traditions have a word for "great teacher"; Chan has a word for "the role that gets tested in every generation, including the current one." That is a different kind of concept, and it is what the three sources, taken together, are pointing at.

If you take one image away, let it be the single sandal. The 祖師 passed through. The trace is real. You cannot find them through it. You can, however, be pointed by it, in your own next ordinary step. That pointing is 祖意. Whether you can actually be pointed — whether the sandal can still point a foot forward — is the only test the tradition has ever cared about.

Canonical EntryAI-generated

定义与核心要义

「祖师」一词,于禅宗典籍中指传承历代法脉之宗师,亦可在当下问答中指可被勘验指认之前人。诸经虽不对其作系统名相解释,却以言句、公案、诗偈、谱系等多种方式呈现其义。

主要阐述

禅宗典籍论「祖师」,并非以传记或义解方式界定,而将其置于机锋问答与日常行履之中加以勘验。

《古尊宿语录》以「从上祖师」并举三世诸佛,重在说法之际学人如何领会,而非建立名相义解;又以「祖师遗下一只履,直至如今觅不得」呈现祖师遗迹虽被传说,却不可循物执取之机锋语气。参礼祖塔、询问祖师言句,乃当前问答中显出见地之勘验,非徒以礼拜为足——僧答「礼塔来」,师即问「祖师道什么」;僧自南华塔头来,师先问「祖师有什么言句」并警「不得错举」;又问「还见祖师么」,以「用见作什么」「既去无过,见有什么过」往复,将「见祖师」转为现前机语。祖师亦可于禅堂中当下指出:师问「堂中还有祖师也无」,僧答「有」,师便说「唤来与老僧洗脚」,以日常动作追问所答能否落实。又有以识与不识、时劫有无发问者,如「还有不识祖师底人么」「什么劫中无祖师」,皆不作系统义理申释,保留其机锋形式。「祖意」一词亦用以表述祖师宗旨:偈云「昔人曾指出,今古道传馨。祖意凭君悟」,把前人指示、古今传承与学人自悟相连;又借披紫衣问「祖师挂什么衣」,以「终日着衣未曾挂着一缕丝」开示祖师之行不滞于衣相。

《景德传灯录》则以史传方式具体建构祖师谱系,自迦叶至般若多罗列有二十七祖,若叙狮子尊者傍出达磨达四世则二十二人,总四十九祖;从七佛至此璨大师不括横枝凡三十七世。当时已流传「祖图」,记五十余祖之名,然「支派差殊宗族不定,或但有空名者」,可见祖师谱系在当时尚未完全定案。昙曜于后魏佛法沦替之际,以素绢单录诸祖名字藏衣领中隐于岩穴三十五载,至文成帝时方重新整理,「集诸沙门重议结集」,可见祖师名号流传历经危难方得延续。全书以「法嗣」为纲,大量罗列祖师门下传承法脉——如「南阳慧忠国师法嗣五人」、「洛阳荷泽神会大师法嗣一十八人」,每一位禅师皆可为某一代祖师。其传承以心法为本,可大师付法云「吾若言有法与人,即为诳汝,但且随方解缚假名三昧」,强调祖师传承非文字法义,乃直指心性之解缚之教。

《明觉禅师语录》以机锋提唱呈现祖师之参究境地。核心表述为「祖师不到处,时人知有;时人不知处,过在祖师」,并追问「作么生辩」而代答「不得春风华不开」,以辨验式问答引人当下举示;后文再由「个个道我会」「会即且致,作么生举」推进,表明口称领会仍须落实于实际举呈。诗偈中亦以祖师为转折归趣所在:「山万重兮水万枝,堆青流碧冷便宜。算来免得生遥恨,不在诗情在祖师」;另有写远别海涯、杪秋挂帆、岛树寒叶、浪天行旅而问「人谁访祖师」,末以「早晚归林下,千徒不共知」作结,呈现寻访祖师之情景与多人追随而未能共知并置之意象。

不同部类与译本的表述差异

三部典籍同用「祖师」一词,然表述各有侧重。

《古尊宿语录》主要称谓为「祖师」,亦见「从上祖师」、「祖意」,未另以同义名目加以解释;所引「祖师遗下一只履」之说以机锋语气论祖师遗迹,与史传部之具体列名方式不同。

《景德传灯录》以「祖师」为主,并用「禅祖」、「祖」并称传承者,如「我师禅祖中当得为第八」;另以「祖师西来意」一语作为禅门公案核心提问,如「如何是祖师西来意。师曰。有力者负之而趋」;又以「法嗣」一词总称祖师门下传承法脉,与《古尊宿语录》之口语勘验方式有别。

《明觉禅师语录》所用术语即「祖师」,未改称其他同义名目,亦未列举某位祖师姓名或历代传承次第;其独特表述为「祖师不到处」、「时人知有」、「时人不知处」、「过在祖师」等组语,以问答结构呈现祖师之义,与另二经之表述方式皆不相同。

另需辨明者,「祖师」与「宗师」并非同义。《明觉禅师语录》序文自称「有愧于宗师」,此「宗师」为圆应对所参承师者之尊称;正文「祖师」则出于机锋提唱及诗偈之中。《古尊宿语录》中此二词亦应分别保留,不可径行替换。

与相关概念的关系

「祖师」与「宗师」:两者并非同义。前者指传承法脉之历代宗师或当下可被勘验之禅门榜样;后者为对所参承师者之尊称,如《明觉禅师语录》序文中「有愧于宗师」即属此用法。

「祖师」与「禅祖」、「祖」:三者皆可指传承禅法之历代宗师,《景德传灯录》中「我师禅祖中当得为第八」即以「禅祖」与「祖」并称传承者。

「祖师」与「法嗣」:祖师为传承法脉之宗师,法嗣为祖师门下之传承弟子,如《景德传灯录》以「法嗣」为纲大量罗列祖师门下传承。

「祖师」与「从上祖师」、「祖意」:《古尊宿语录》以「从上祖师」并举三世诸佛,「祖意」则指祖师之宗旨。

「祖师」与「祖图」:《景德传灯录》载当时已流传「祖图」,记五十余祖之名,可视为祖师谱系之图录形式。

代表性原文引用

《古尊宿语录》:「从上祖师.三世诸佛说法。」「祖师遗下一只履,直至如今觅不得。」师问「祖师有什么言句」,并警「不得错举」。僧答「礼塔来」,师即问「祖师道什么」。「还见祖师么」「既去无过,见有什么过」。「堂中还有祖师也无」「唤来与老僧洗脚」。「还有不识祖师底人么」「什么劫中无祖师」。「昔人曾指出,今古道传馨。祖意凭君悟。」「祖师挂什么衣」「终日着衣未曾挂着一缕丝」。

《景德传灯录》:「自迦叶至般若多罗,有二十七祖。若叙狮子尊者傍出达磨达四世,二十二人,总有四十九祖。若从七佛至此璨大师,不括横枝凡三十七世。」「以素绢单录,得诸祖名字……藏衣领中隐于岩穴,经三十五载。」「我师禅祖中当得为第八。」「如何是祖师西来意。师曰。有力者负之而趋。」「南阳慧忠国师法嗣五人」、「洛阳荷泽神会大师法嗣一十八人」。「吾若言有法与人,即为诳汝。但且随方解缚假名三昧。」

《明觉禅师语录》:「祖师不到处,时人知有;时人不知处,过在祖师。作么生辩。不得春风华不开。」「个个道我会」「会即且致,作么生举」。「山万重兮水万枝,堆青流碧冷便宜。算来免得生遥恨,不在诗情在祖师。」「人谁访祖师」「早晚归林下,千徒不共知」。

By SutraAI-generated

明覺禪師語錄T1996 · 諸宗部

《明覺禪師語錄》論「祖師」,主要不是以名目訓詁或世系敘述加以界定,而是在提唱、代語與詩偈之中,呈現祖師及其禪法宗旨所關涉的參究境地。其核心表述為:「祖師不到處,時人知有;時人不知處,過在祖師。」此語將「祖師不到處」與「時人知有」相對,又將「時人不知處」歸為「過在祖師」,隨即追問「作麼生辯」,並代答「不得春風華不開」。語錄因此不對「祖師不到處」作義理解說,而以辨驗式問答引人當下舉示;後文再由「箇箇道我會」「會即且致,作麼生舉」推進,表明口稱領會仍須落實於實際舉呈。 本經使用的主要術語形式即「祖師」,未在所引原文中改稱其他同義名目,也未列舉某位祖師的姓名或歷代傳承次第。與此相近而不可混同的用詞是「宗師」:序文自稱多聞未益,因而「有愧於宗師」,此處是圓應對所參承師者的尊稱;正文的「祖師」則出現在機鋒提唱及詩偈中。故記錄本經用語時,應分別保留「祖師」與「宗師」,不可逕將二詞替換。 「祖師」在本經中的一個獨特側面,是與「不到處」「知有」「不知處」「過」等語連用。這組說法並未直接陳述祖師的身分、德號或傳法譜系,而把祖師置於「時人」能否知有、能否辨明的問答結構中;「不得春風華不開」則是原文所給代語,不宜越出語境另作義理裁定。 另有詩偈云:「山萬重兮水萬枝,堆青流碧冷便宜。算來免得生遙恨,不在詩情在祖師。」此處以山水與詩情鋪陳,末句卻將著眼處明確落在「祖師」,顯示「祖師」可成為詩偈轉折與歸趣所在;原文只以「不在……在……」標示重心,並未進一步說明其所指。 又一偈寫遠別海涯、杪秋掛帆、島樹寒葉與浪天行旅,繼而問「人誰訪祖師」,末以「早晚歸林下,千徒不共知」作結。此處的「祖師」與「訪」字相連,呈現尋訪祖師的情境;同時以「千徒不共知」收束,使多人追隨與未能共知並置。然偈中不明言祖師為誰,也不交代所訪之法,故只能據文記錄其尋訪、歸林及不共知的意象,不宜外推為特定人物或傳承事件。

景德傳燈錄T2076 · 史傳部

《景德傳燈錄》作為禪宗史傳部的重要典籍,以「祖師」一詞指稱歷代傳承禪法宗旨的宗師,其論述呈現以下幾個層面:

一、祖師譜系的具體建構。書中明列祖師傳承的代數與世系。李常與西域三藏犍那問答中提及「自迦葉至般若多羅,有二十七祖。若敘獅子尊者傍出達磨達四世,二十二人,總有四十九祖。若從七佛至此璨大師,不括橫枝凡三十七世」,清楚勾勒出從印度初祖迦葉到中國禪宗祖師的法脈傳承。此段亦揭示當時已流傳「祖圖」,記錄五十餘祖之名,但「支派差殊宗族不定,或但有空名者」,可見祖師譜系在當時尚未完全定案。

二、祖師譜系的保存與考證。針對祖師傳承的可靠性,書中記載曇曜於「後魏初佛法淪替」之際,「以素絹單錄,得諸祖名字……藏衣領中隱于巖穴,經三十五載」,至文成帝即位法門中興時方重新整理,「集諸沙門重議結集」,顯示祖師名號的流傳歷經危難方得以延續。

三、祖師之稱的多樣表述。書中以「禪祖」、「祖」並稱傳承者,如「我師禪祖中當得為第八」,並以「祖師西來意」一語作為禪門公案的核心提問,如「如何是祖師西來意。師曰。有力者負之而趨」之問答。「祖師」與「禪祖」並用,皆指傳承禪法之歷代宗師。

四、祖師法嗣的傳承網絡。全書以「法嗣」為綱,大量羅列祖師門下的傳承法脈,如「南陽慧忠國師法嗣五人」、「洛陽荷澤神會大師法嗣一十八人」等,顯示祖師概念在本書中不只是少數幾位宗師,而是綿延不絕的法脈網絡,每一位禪師皆可為某一代祖師,其門下弟子即為法嗣。

五、祖師之傳承以心法為本。書中記載可大師付法之語:「吾若言有法與人,即為誑汝。但且隨方解縛假名三昧」,強調祖師傳承非文字法義,而是直指心性的解縛之教。

古尊宿語錄X1315 · 禪宗語錄部

《古尊宿語錄》論「祖師」,主要不以傳記式敘述界定歷代宗師,而將「祖師」置於禪門問答、機鋒與日常行履之中,藉其言句、遺蹟及宗旨,考驗學人能否當下領會。

一、經中最常用的稱謂就是「祖師」,亦見「從上祖師」及「祖意」;並未在所引文字中另用其他同義名目加以解釋。「從上祖師.三世諸佛說法」一語,把歷來祖師與三世諸佛並舉;所問卻是山河大地草木何以「不省去」,重點落在說法之際學人如何領會,而非建立名相義解。

二、祖師被視為留有宗旨、言句與蹤跡者。經文說「祖師遺下一隻履,直至如今覓不得」,以「一隻履」與「覓不得」呈現祖師遺蹟雖被傳說,卻不可循物執取的機鋒語氣。 又有僧自南華塔頭來,師先問「祖師有什麼言句」,並警告「不得錯舉」,顯示祖師言句在禪門中須被準確舉唱,且能否承當其意,是問答所檢驗之事。

三、經中屢以參禮祖塔、詢問祖師作為勘驗。僧答「禮塔來」,師即問「祖師道什麼」;僧反問後,師評為「將謂是箇靈利漢」,表明到祖塔禮拜並不足以證成領會,仍須在當前問答中顯出見地。 同樣,僧說從南華塔頭來,師問「還見祖師麼」;雙方又以「用見作什麼」「既去無過,見有什麼過」往復,將「見祖師」轉成現前機語,而非單指形貌上的看見。

四、「祖師」也可用來指禪堂中可被當下指出者。師問僧「堂中還有祖師也無」,僧答「有」,師便說「喚來與老僧洗脚」。此則不鋪陳祖師身分,反以洗脚這一日常動作追問「有祖師」之答能否落實,呈現本經把祖師之義置於具體應對中的獨特側面。

五、經文亦以識與不識、時劫有無來發問。師問「還有不識祖師底人麼」,代語答「仁義道中」,又答「不患」;原文只錄問答,未進一步申釋何謂識祖師,因此宜保留其機鋒形式,不另作義理推演。 又問「既知來處,什麼劫中無祖師」,代語答「某甲今年不著便」,以反詰表示祖師並非只繫於某一特定劫時,但經文仍不作系統說明。

六、祖師宗旨又以「祖意」表述。偈云「昔人曾指出,今古道傳馨。祖意憑君悟」,把昔人之指示、古今之傳承與學人自悟相連;祖意雖由前人指出並流傳,領悟仍須由當人承擔。 此外,師披紫衣時問「祖師挂什麼衣」,隨即以「終日著衣未曾挂著一縷絲」開示,顯示祖師之行不由衣相判定,並借披衣一事呈現不滯於所著之意。

Related