Dharma · AI Companion

vāc

语(vāc),指口所发之语言行为,与身、心并称三业之一。AI-generated

Type ViewDifficulty IntroductoryTruth-level span I · dependent arisingInitial AI estimate · evolves with use

A Contemporary ReadingAI-generated

What this concept is saying

語 (vāc, “speech” or “verbal expression”) is language enacted: meaningful expression issuing through the mouth and, in the ethical context, one of the three doors of karma—身業 (kāya-karman, bodily karma), 語業 (vāk-karman, verbal karma), and 意業 (mano-karman, mental karma).

“Speech” is the closest English translation, but it can sound too static. 語 is not merely a vocabulary, a language system, or a collection of sentences. It is the event of speaking: intention taking audible, intelligible form and entering a shared world. A thought may remain private; once spoken, it can become another person’s fear, confidence, resentment, understanding, or resolve. Speech is therefore a bridge by which inner formation becomes interpersonal karma.

The classical sources approach this phenomenon at several distinct levels. They should not be collapsed into one definition.

First, in the narrow Abhidharma sense, 語 is not every sound made by the mouth. The *Abhidharmakośabhāṣya* states:

“Not every sound is called speech. Only a sound through which a meaning can be understood is called speech.”

A cough, a yawn, or an accidental cry is vocal sound, but not necessarily 語 in this technical sense. Speech must be conventionally capable of conveying a determinate meaning. The hearer need not agree with it, and the statement need not be true; even a lie is speech precisely because its false meaning can be understood.

This also explains why 語 must not simply be equated with 名、句、文—“names,” “sentences,” and “phonemic or syllabic elements.” These are closely related dimensions of linguistic articulation, but the *Abhidharmakośabhāṣya* treats their relation as a genuine philosophical problem rather than as an identity. Speech is the meaningful vocal performance; names designate, sentences complete a meaning, and verbal elements support their articulation. The treatise examines and rejects several overly simple theories—for example, that speech straightforwardly “produces” a separately existing name or merely “reveals” one. Language works through distinctions and conventions, but its functioning should not be explained by casually turning every linguistic level into the same entity.

Second, the *Yogācārabhūmi* uses 言論 (verbal discourse) as a comprehensive category embracing 言說、言音、言詞: acts of saying, vocal or phonetic form, and linguistic expressions. This is broader than the narrow question “Which sound counts as speech?” It attends to the whole movement from vocal activity to articulated language and meaningful wording.

The same treatise also places 語 beside walking, standing, sitting, lying down, waking, and 默 (silence) as an activity in which 正知 (clear comprehension) must be maintained. Here speech includes:

- reciting teachings clearly; - explaining and opening them to others; - conversing with wise fellow practitioners or virtuous laypeople; - offering encouragement and appropriate social reassurance; - communicating about legitimate practical needs.

Speech is thus not an unfortunate interruption of “real practice.” Speaking itself is a field of practice. Silence is another field: withdrawing to reflect on teachings already received, cultivating concentration, and developing calm and insight. Neither noise nor muteness is inherently liberating. The question is whether one knows what one is doing, why one is doing it, and what qualities of mind are being strengthened.

Third, 語 is 語業—verbal karma. “Karma” here does not mean a supernatural punishment attached to forbidden words. It means intentional activity with formative and consequential force. Speech displays a present intention, conditions the speaker’s habits, alters relationships, and may leave effects that long outlast the sound.

The *Abhidharmakośabhāṣya* says that bodily and verbal karma possess both 表 (manifest or communicative expression) and 無表 (non-manifest efficacy or restraint). This is a technical Abhidharma position, not simply another word for a spoken sentence. The manifest aspect is the speech-event through which intention becomes perceptible. The non-manifest aspect, in the relevant scholastic account, is a continuing karmic efficacy associated with commitments, restraints, and completed deeds even when no fresh sound is being produced. One should not reduce 無表業 to “subconscious memory,” “body language,” or an invisible mystical vibration; those are modern substitutions for a specific doctrinal category.

The treatise’s analysis of false speech makes the ethical structure particularly clear. It describes 虛誑語 (deceptive or false speech) as speaking from a defiled intention while holding a conception different from what one presents. It then distinguishes stages in the act. The preliminary words are preparation; the completed karmic course is established when the person being deceived understands the intended false meaning. This is not saying that a failed attempt to lie is harmless. It distinguishes an attempted deed from a fully completed karmic pathway. Intention matters, expression matters, and successful communication matters.

This prevents two opposite mistakes. A technically false statement uttered through innocent confusion is not ethically identical to deliberate deception. Conversely, a speaker cannot wash away deliberate manipulation by saying, “Those were only words.” Words are precisely the medium through which the intention was made effective.

Fourth, the *Mahāratnakūṭa Sūtra* considers 語 as one member of the three karmic doors and lists different combinations of purity and impurity: bodily conduct may be pure while speech and mind are impure; body and speech may appear pure while the mind is impure; the mind may be pure while body and speech are not; speech and mind may be pure while bodily conduct is not. In this passage 語 is an abbreviation for 語業.

The point is subtle. A person is not morally described by a single flat label. Someone may refrain from physical harm yet use language cruelly. Someone may speak politely while inwardly cultivating contempt. Someone may have a benevolent intention yet speak carelessly and cause avoidable injury. Someone else may use imperfect words while sincerely trying to protect another person. The three doors interact, but they do not automatically coincide.

This is not permission to excuse harmful speech with “my heart was in the right place.” It is an instruction to examine the whole causal configuration. Liberation requires more than managing appearances, and more than admiring one’s intentions. Body, speech, and mind must each be trained.

That places 語 within the path of liberation. Speech stands between inner intention and the social world. It is conditioned by perception, feeling, craving, aversion, views, and habits; once expressed, it reinforces some of those very conditions. Unexamined speech therefore helps perpetuate 流轉 (cyclic continuation): irritation becomes accusation, accusation produces defensiveness, defensiveness confirms one’s hostile story, and the story prepares the next accusation. Carefully understood speech can interrupt that circuit.

This is why 正語 (samyag-vāc, Right Speech) belongs to the Noble Eightfold Path. Its purpose is not merely to make a practitioner pleasant or respectable. It disciplines the outward channel through which greed, hatred, and delusion recruit other people into their momentum. Traditionally, Right Speech is explained through abstention from false, divisive, harsh, and frivolous or heedless speech. Its positive horizon is language oriented toward truthfulness, reconciliation, kindness, meaningfulness, and an appropriate occasion. Silence may sometimes serve these qualities, but silence is not automatically their fulfilment.

Finally, Chan sources add a different but complementary warning. The *Recorded Sayings of Chan Master Dongshan Liangjie of Ruizhou* says:

“Only after embodying the matter beyond Buddha does one have some share in speech.”

It also speaks paradoxically of not hearing while speech is occurring, hearing when there is no speech, and of the “correct question and correct answer” as “not spoken from the mouth.” These sayings should not be used to erase the earlier ethical teaching, as though awakening made truthfulness irrelevant. Their target is attachment to verbal formulations. A person may repeat flawless Buddhist language without seeing what it indicates. Words can point, disclose, question, and transform, but no formulation can substitute for realization.

Thus the traditions hold two truths together: speech matters enormously, and words are not the reality to which wise speech points.

A walkthrough in everyday life

Imagine an exhausted commuter standing in a crowded train after a difficult day. A message arrives from a partner:

“You forgot again. Do I matter to you at all?”

The commuter feels a surge of heat and types:

“I told you I was busy. You never listen. Stop making everything about you.”

Before sending it, we can walk through the event using the classical distinctions.

1. Mental karma is not yet verbal karma.

The flash of anger, the remembered argument, the wish to defend oneself, and the intention to strike back belong primarily to the mental door. They condition possible speech, but they are not identical with speech. Thinking “You never listen” is not yet an audible utterance in the narrow classical definition.

This is where people often map the teaching incorrectly: they call every thought “inner speech” and then conclude that mental and verbal karma are the same. Ordinary psychology may use “inner speech” as a useful term, but the three-karmas analysis preserves a distinction between mental activity and verbal expression.

2. The phone’s alert is sound, but not the commuter’s 語.

The notification tone is audible, yet it is not meaningful speech issuing from the commuter. If a voice assistant reads the partner’s message aloud, meaningful language is heard, but the questions “Whose verbal action is this?” and “Which intention is karmically operative?” become more complicated. The device generates sound; the partner composed the message; a platform transmitted it; the commuter interprets it. Classical categories can illuminate these layers, but they should not be forced into a claim that ancient texts explicitly classified smartphones or synthetic voices.

3. Typing is not literally mouth-produced vocal sound.

The narrow definition in the supplied Abhidharma material concerns meaningful sound. A typed message is therefore not identical to spoken 語 under that strictly vocal definition. Yet in contemporary ethical application, messaging performs many of the same communicative functions as verbal karma: an intention becomes articulated language, another person understands it, and relationships are changed by it.

Calling a text “speech” is thus a reasoned functional extension, not a quotation from the classical definition. This distinction matters. We preserve the text’s boundary while still allowing its ethical analysis to address algorithmic and digital life.

4. Names, sentence structure, and the act of communicating are related but not identical.

“You,” “busy,” and “listen” are linguistic units. “You never listen” is a sentence with an intelligible claim. But the verbal act includes more than dictionary meanings. In this context it performs blame, exaggeration, self-protection, and perhaps retaliation. Knowing the vocabulary does not yet disclose the karma of its use.

5. Intention must be examined precisely.

Does the commuter genuinely believe the partner never listens? Is “never” an impulsive exaggeration? Is the aim to report a fact, end the conversation, punish the partner, or compel reassurance?

If the commuter knows the claim is false and deliberately presents it as true to manipulate the partner, the structure approaches deceptive speech. If the commuter sincerely but mistakenly believes it, the factual error remains important, but it is not the same intentional configuration. If the commuter chooses a true fact specifically to humiliate—“Everyone finds you exhausting”—truth alone does not purify the act. Speech can be factually accurate and still harsh, divisive, mistimed, or motivated by ill will.

6. Completion involves reception.

Suppose the message is sent but the network fails. The intention and attempt have already occurred, so this is not ethically nothing. Yet the *Abhidharmakośabhāṣya* distinguishes preparation from a completed course of false speech: completion occurs when the person targeted understands the deceptive meaning.

If the message is delivered and the partner reads it as blame, the act has entered another stream of experience. The recipient may feel shame, become defensive, or answer with a sharper accusation. The speech has now helped build the next conditions in the cycle.

7. Body, speech, and mind may differ.

The commuter might delete the message and send, “I’m overwhelmed and defensive. I did forget, and I’m sorry. Can we talk when I get home?”

The body is still tense. The mind may still contain anger. The message is more truthful and less injurious, but it does not prove that the mind has become pure. This resembles the *Mahāratnakūṭa*’s refusal to assume that purity in one karmic door guarantees purity in the others.

Conversely, suppose the commuter feels genuine care but sends a rushed, ambiguous “Fine.” A relatively benevolent intention does not ensure skilful expression. The other person receives language, not telepathic access to the speaker’s good heart. Training must include both inward motivation and communicative form.

8. Clear comprehension asks what speech is for at this moment.

The *Yogācārabhūmi* treats speaking and silence as activities to be known clearly. The commuter can ask:

- What am I about to communicate? - What state of mind is propelling it? - Is the wording accurate? - Is this the right medium and time? - Will speaking now clarify, reconcile, encourage, or meet a legitimate need? - Would temporary silence create space for reflection, or would it function as punishment and avoidance?

Pausing is not repression if the pause is used to understand and choose. Nor is immediate “honesty” automatically courageous. Sometimes “I’m just being honest” means “I want my present anger to speak before wisdom can examine it.”

9. Digital speech has an extended afterlife.

A spoken insult may fade acoustically, though not psychologically. A post, screenshot, forwarded voice message, or generated clip can be copied and recirculated. Algorithms may amplify language because it provokes engagement, not because it is true or beneficial. This intensifies the practical importance of verbal karma: one impulsive expression can repeatedly encounter new recipients.

The analogy has limits. An algorithm has no established status in Buddhist doctrine as a morally responsible person, and current science and philosophy do not offer a settled account of machine intention or consciousness. We should not claim that AI “has karma” merely because it generates sentences. The clearer question is how human intentions, institutional choices, training data, interface design, and acts of sharing contribute to communicative consequences.

10. Chan introduces one final inspection.

Even the improved message—“I’m overwhelmed, I forgot, and I’m sorry”—can become a script performed without contact with experience. One can learn therapeutic phrases and use them mechanically, strategically, or self-righteously. Dongshan’s warning asks whether words arise from actual seeing or merely imitate the sound of insight.

This does not mean waiting for enlightenment before apologizing. It means not mistaking the correct formula for complete understanding. A living apology is supported by acknowledgment, changed conduct, and the willingness to hear what the other person says. The words point beyond themselves into a transformed relationship.

Why contemporary people need it

Contemporary life treats speech as nearly weightless and yet builds entire economies around it. Messages are sent in seconds, opinions become public before they mature, and platforms reward speed, certainty, outrage, and repetition. Language feels disposable at the moment of production but can become permanent at the moment of circulation.

The Buddhist analysis restores weight without turning speech into a purity contest.

It first teaches that communication is conduct. Posting, commenting, instructing an AI system, forwarding a rumour, “just asking questions,” or withholding a crucial qualification are not merely transfers of information. They are ways of shaping attention and relationship. Speech trains the speaker as it affects the hearer. Each repetition makes certain responses easier to repeat: sarcasm, evasiveness, careful truthfulness, patient explanation, or reconciliation.

Second, the three-door framework protects us from performative virtue. A person can maintain polished public language while privately feeding hatred. Another may cultivate compassionate intentions while refusing to learn how their words actually land. Ethical development cannot be reduced either to visible etiquette or to invisible sincerity. Mind, expression, and embodied conduct must be brought into increasing coherence.

Third, the distinction between sound, meaning, and linguistic form helps us resist verbal enchantment. A slogan may be memorable without being true. Technical vocabulary may create an impression of knowledge without conveying understanding. AI-generated prose may be fluent while containing fabricated facts or citations. Fluency is evidence of linguistic organization, not necessarily of wisdom, sincerity, or realization.

This is one place where contemporary cognitive science and language technology offer useful comparisons. Research can study language production, emotional arousal, social influence, and the brain’s processing of speech. AI systems demonstrate dramatically that coherent verbal patterns can be produced without thereby settling questions about understanding or consciousness. But none of this “proves Buddhism.” Scientific models investigate measurable processes under specific assumptions; Buddhist teachings about karma and liberation address ethical cultivation, suffering, and release. The domains can illuminate one another without being declared identical.

Fourth, training speech reveals the mind with unusual clarity. Meditation may leave us with a flattering image of our calmness until someone contradicts us. Then tone, exaggeration, interruption, strategic silence, and the hunger to win show exactly where attachment remains. Speech is a diagnostic surface on which subtle greed, aversion, fear, and conceit become observable.

Fifth, speech is indispensable to the path even though liberation is not a verbal achievement. Teachings are heard, recited, questioned, explained, and discussed. Practitioners encourage one another. Confusion is corrected through language. Yet the Chan warning remains: a map is not arrival, and eloquence about non-attachment can itself become an attachment.

Wise speech therefore serves awakening in at least three ways. It restrains the creation of fresh harm; it cultivates mental qualities compatible with clarity and compassion; and it communicates teachings that point beyond conceptual possession. At maturity, one does not despise words or worship them. One uses them responsibly and knows their limit.

Common misunderstandings and clarifications

1. “語 just means words or language.”

Not exactly. In its narrow technical sense, it is meaningful vocal expression—sound through which a conventionally established meaning can be understood. In broader Yogācāra usage, verbal discourse includes saying, vocal form, and linguistic expression. In the karmic context, it is intentional verbal conduct. These levels overlap but should not be flattened.

2. “Every sound from the mouth is speech.”

No. The *Abhidharmakośabhāṣya* explicitly excludes the idea that all sound is 語. A cough or involuntary cry is not automatically meaningful speech. Context can change the case: a deliberately coded cough may function communicatively because a shared meaning has been assigned to it.

3. “Speech karma is just the sentence itself.”

No. The ethical act includes intention, meaningful expression, context, and—in the analysis of completed deceptive speech—the recipient’s understanding. Identical words can function as comfort, irony, threat, quotation, deception, or instruction.

4. “If a statement is true, it is Right Speech.”

Truthfulness is necessary but not sufficient. A true statement can be weaponized, disclosed at a destructive time, used to divide people, or delivered with deliberate cruelty. Right Speech concerns the whole orientation of the act, not factual correctness alone.

5. “If my intention was good, my speech was pure.”

Good intention matters greatly, but it does not make communicative skill unnecessary. The *Mahāratnakūṭa*’s permutations of pure and impure body, speech, and mind warn against assuming automatic alignment. A compassionate aim can be expressed carelessly; honest feedback from those affected is therefore part of practice.

6. “If the hearer misunderstood me, I bear no responsibility.”

That is too simple. Sometimes misunderstanding is unforeseeable; sometimes the speaker relies on ambiguity, insinuation, coded hostility, or plausible deniability. Buddhist analysis encourages an honest examination of intention and causal contribution, not either total self-condemnation or automatic self-exoneration.

7. “An unsuccessful lie produces no karma.”

The cited Abhidharma analysis distinguishes an attempted deception from the completed karmic course established when the target understands the false meaning. It does not turn preparation into innocence. The attempt already manifests and strengthens a deceptive intention, even if a technical condition for completion is absent.

8. “無表業 means an invisible message or subconscious body language.”

No. It is a school-specific Abhidharma category concerning non-manifest karmic efficacy, especially relevant to restraints and commitments. It should not be casually translated into a modern psychological or physical mechanism.

9. “Texting and posting are literally the same as the classical vocal definition of 語.”

Not literally. The narrow definition is sound-based. Treating digital writing as verbal conduct is a contemporary functional extension based on its role in conveying intention and meaning. Stating the extension openly is more faithful than silently rewriting the classical boundary.

10. “Silence is always spiritually superior to speech.”

No. The *Yogācārabhūmi* treats both 語 and 默 as modes of practice requiring clear comprehension. Silence may support reflection and meditation; it may also conceal fear, indifference, passive aggression, or refusal to protect someone. Speech may be distracting; it may also teach, reconcile, encourage, confess, or prevent harm.

11. “Chan says true speech does not come from the mouth, so ethical rules about speech are merely provisional.”

That conclusion confuses two levels of instruction. Dongshan exposes attachment to words and imitation of realization. He does not provide a licence for lying or cruelty. Transcending fixation on language is not the same as becoming indifferent to the suffering language can cause.

12. “A person who speaks beautifully must understand the Dharma.”

Eloquence can transmit understanding, but it can also imitate it. Buddhist vocabulary, therapeutic fluency, academic sophistication, and AI-generated polish can all produce convincing surfaces. The deeper test is whether speech arises with truthful seeing, reduces greed and hostility, withstands examination, and is accompanied by transformed conduct.

語 is therefore neither a minor etiquette topic nor merely a linguistic category. It is the threshold where an inward world becomes shared causality. To train speech is to notice the instant in which feeling becomes a story, the story becomes an intention, the intention finds words, and the words begin living in other minds.

The liberating task is not to become silent forever or verbally flawless. It is to understand this threshold so intimately that delusion no longer crosses it unnoticed. Then speech can become truthful without cruelty, compassionate without manipulation, precise without vanity, and silent without withdrawal. And when words have done what words can do, one can let them fall away without mistaking their absence for awakening.

Canonical EntryAI-generated

定义与核心要义

语(vāc),梵语 vāc 之译,意指口所发之语言行为,为佛教术语中三业(身业、语业、意业)之一,与身、心并称。语之概念在佛教经论中有广狭诸义:狭义指由口所发、能令了义之音声;广义则涵盖一切言说、言音、言词之活动,与名、句、文等概念相涉而又互别。

主要阐述

《大宝积经》将「语」置于如来第六力(知处非处智力)之遍知境界下,与「身」、「心」对举并列,以三业净秽交络为切入点。经中列举种种净秽交错之组合:身业清净而语业、心业不清净;语业与身业清净而心不清净;唯心清净而身语不清净;语业与心清净而身业不净等,总摄三业净秽流转与寂灭,皆为如来遍智所了。

《阿毘达磨俱舍论》以「语」为能诠定量之音声所成之业行。论中先厘清语与名、句、文的分际:三者虽以「声」为体而属色法,其自性却非即是语;唯能令了义之音声方称为语,如古者共以瞿声立于九义。又破斥「语生名」、「语显名」、「语生文、文复生名」等执。在业的范畴中,以「身语二业,俱表无表性」明示语业与身业同具表、无表二性。于虚诳语业道之成立,以「染异想发言,解义虚诳语」立名,辨成业道之时:所诳者解所说义时,与最后字俱生之表声及无表业方成业道,前行诸字皆是加行。又以颂明身、语与能发之业并列,示语为能招福果之行;复论神通中「一化主语时,诸所化皆语」之语相。

《瑜伽师地论》从三个层面展开论述:首先,以「言论者,谓一切言说、言音、言词,是名言论」作出定义性表述,涵盖口头发声、语音形态与语言单位三个面向;其次,在「正知住」修行脉络中,将「语」列为与行、住、坐、卧等并列的威仪之一,与「默」对扬,涵盖读诵经典、开示教化与日常酬应;「默」则是于空闲之处思惟已受之法、修习禅定;第三,在论「文」之体性时,「语」被列为六种「文」之一,与名身、句身、字身、行相、机请并举。

《瑞州洞山良价禅师语录》从禅宗立场对「语」作出深刻反省,强调真正的「语」须建立在见性基础之上。师云:「体得佛向上事,方有些子语话分。」 又以「语话时,阇黎不闻」与「不语话时即闻」之对比,点出语之真实在于言语背后所指向的境地。复以「如何是正问正答?不从口里道」明真正的问答并非透过口舌而发,将「语」从口语拉升至悟境层面。

不同部类与译本的表述差异

各部类经论对「语」之用语有所不同:

- 《大宝积经》以「语」字直接替代「语业」,未完整使用「语业」一词,此为本经特有的简称用法,如「净修于身业,非语非心业清净」、「语净及身净」、「语净及心净」、「身语二业非清净」中之「语」皆为语业之略称。此一以「语」代「语业」的表述方式,与其他经典独立使用「语业」一词之惯例不同。

- 《阿毘达磨俱舍论》则使用完整之「语业」概念,并将其与名、句、文等概念作精细区辨。

- 《瑜伽师地论》使用「语」与「言论」两个相关而各有侧重的术语——「语」在威仪与文法两脉络中皆属较广的活动范畴,「言论」则是对此范畴所作的定义性总名,以「言说—言音—言词」三层展开。

- 《瑞州洞山良价禅师语录》则使用「语话」一词,与其他经论中常见的「语言」、「言说」等用语有所不同,侧重于言谈活动本身的动态性质,而非静态的语言体系。

与相关概念的关系

语与身、心并称三业,三者构成有情行为之三大门类。语又与名、句、文密切相关:语为能诠定量之音声所成之业行,与名句文相即而不一,名、句、文三者以声为体而属色法,却非即是语。语与「默」形成对扬,分别指向语言行为与静默思惟之两种修行威仪。在文法施设上,「语」为六种「文」之一,与名身、句身、字身等并举。在禅宗语境中,语又与「悟」紧密相联,真正的语须建立在见性基础之上,指向超越口舌的不言之地。

代表性原文引用

- 《大宝积经》:「净修于身业,非语非心业清净。」「或身语心净不净,诸行流转及寂灭,遍智见者如实知。」

- 《阿毘达磨俱舍论》:「非但音声皆称为语,要由此故义可了知,如是音声方称语故。」「身语二业,俱表无表性。」「由此舍名施……身语及能发,此招大富果。」「一化主语时,诸所化皆语。」

- 《瑜伽师地论》:「言论者,谓一切言说、言音、言词,是名言论。」

- 《瑞州洞山良价禅师语录》:「体得佛向上事,方有些子语话分。」「语话时,阇黎不闻。」「不语话时即闻。」「不从口里道。」

By SutraAI-generated

大寶積經T0310 · 寶積部

《大寶積經》(唐菩提流志譯,寶積部)中,「語」一詞出現於論述如來第六力(知處非處智力、遍智見者如實知)的偈頌段落,與「身」、「心」並列,作為三業(身業、語業、心業)構成要素之一而呈現。

從術語形式觀察,本經在該段中並未完整使用「語業」一詞,而是直接以「語」字替代語業之義,此為本經特有的簡稱用法。諸如「淨修於身業,非語非心業清淨」、「語淨及身淨,而彼心體非清淨」、「語淨及心淨,而彼身業未甞淨」、「身語二業非清淨」,乃至總括句「或身語心淨不淨」中之「語」,皆為語業的略稱。此一以「語」代「語業」的表述方式,與其他經典獨立使用「語業」一詞之慣例不同,宜於多譯名歸一處明確標示。

本經論述「語」的核心側面,在於將其置於如來遍知境界之下,說明佛陀如實了知有情於語業上的種種差別相狀。其論述結構以三業(身、語、心)為綱,逐一列舉淨與不淨之種種交錯組合,以顯諸行相貌之繁複:

一、有行者「淨修於身業,非語非心業清淨」——身業清淨而語業、心業不清淨。 二、有行者「語淨及身淨,而彼心體非清淨」——語業與身業清淨,然心不清淨。 三、有行者「內心常清淨,身語二業非清淨」——唯心清淨,身業、語業皆不清淨。 四、有行者「語淨及心淨,而彼身業未甞淨」——語業與心清淨,唯身業不淨。 五、總收一切,「或身語心淨不淨,諸行流轉及寂滅,遍智見者如實知」,三業之淨穢流轉與寂滅,皆為如來第六力之所遍了。

由此可見,本經並非孤立地詮說語言行為本身的體性或作業過程,而是將「語」作為三業之一,與「身」、「心」對舉並列,側重呈現其與其他二業在清淨與否上的種種交錯差別,從而凸顯有情諸行之繁複與參差,最終統攝於如來遍智所知之下。此種以三業淨穢交絡為切入點的論述路徑,即為本部經典對「語」概念之主要表達方式。

阿毘達磨俱舍論T1558 · 毘曇部

《阿毘達磨俱舍論》以「語」指稱由口所發的言說行為,並將其與名、句、文三類概念相互區辨而立論。

論中首先釐清語與名句文的分際:名、句、文三者雖以「聲」為體而屬於色法,其自性卻非即是語。論云「非但音聲皆稱為語,要由此故義可了知,如是音聲方稱語故」,明確定義「語」為能令了義之音聲——唯有經共立為「能詮定量」的音聲方稱為語,例如古者共以瞿聲立於九義。語發名、名顯義,由此而令了知,並非一切音聲皆得語名。

關於語與名句文的關係,論中破斥「語生名」、「語顯名」、「語生文、文復生名」等執,論及諸念不可聚集、差別已足顯義、何待別名等理。毘婆沙師則別立實有的名等身,攝於心不相應行蘊中,視為「實而非假」之法。

論中亦將「語」結合於業的範疇:以「身語二業,俱表無表性」明語業與身業同具表、無表二性。於虛誑語業道之成立,以「染異想發言,解義虛誑語」之頌立名,繼之辨成業道之時——所誑者解所說義時,與最後字俱生之表聲及無表業方成業道,前行諸字皆是加行;「發言」、「所發言」之語於此屢見。

又以「由此捨名施……身語及能發,此招大富果」之頌,將身、語與能發之業並列,示語為能招福果之行;復有「一化主語時,諸所化皆語」之說,論神通中之語相。整體而言,本論以「語」為能詮定量之音聲所成之業行,與名句文相即而不一,並落實於身語二業、虛誑語等具體業道之辨析。

瑜伽師地論T1579 · 瑜伽部

《瑜伽師地論》論述「語」這一概念時,主要從三個層面展開,並使用了「語」與「言論」兩個相關而各有側重的術語。

第一,本論以「言論」一詞對語言行為作出直接的定義性表述:「言論者,謂一切言說、言音、言詞,是名言論。」 此處將「言論」界定為一切言說、語音與詞語的總稱,三者並列,涵蓋了口頭發聲、語音形態與語言單位三個面向,呈現出從發聲到語義的完整序列。

第二,在「正知住」的修行實踐脈絡中,「語」被列為與行、住、坐、臥、覺寤、默等並列的威儀之一,並與「默」形成對比。文中詳述「語」的具體內容包括:對已受之法「言善通利」,以大音聲讀誦,或為他人廣說開示;於時時間與有智同梵行者或在家的賢善者「語言談論,共相慶慰」,為勸勵及求資具。由此可見,實修中的「語」不限於讀誦經典,更涵蓋開示教化與日常酬應;與之相對,「默」則是於空閑之處思惟已受之法、修習禪定(內住、等住、安住、近住、調伏、寂靜乃至毘鉢舍那)。此一語/默對揚,揭示了本論將語言行為納入修行威儀而予以正知觀察的立場。

第三,在論述「文」之體性時,「語」被列為六種「文」之一,與名身、句身、字身、行相、機請並舉。此處的「語」側重於作為教法表達形式之一的語言層面,與「名」(能指符號)、「句」(完整句義)相區別,構成教法施設的不同面向。

在術語層面值得留意的是:本論以「語」與「言論」分指相關而層次有別的概念——「語」在威儀與文法兩脈絡中皆屬較廣的活動範疇,而「言論」則是對此範疇所作的定義性總名,並以「言說—言音—言詞」三層加以展開;二者並行使用,反映了本論在不同語境下區分語言現象時的細密考量。

Related