A seed left in the storehouse consciousness by the activity of naming, speaking, and conceptual discrimination — the proximate cause (親因緣) by which corresponding dharmas later arise.
一、这个概念在说什么
This concept is the deep grammar of why your world keeps looking the way it does.
To understand 名言熏習 you need three things at once: a theory of how anything "leaves a trace," a theory of where those traces are stored, and a theory of what *specifically* gets traced. The *Mahāyāna-saṃgraha* (《攝大乘論》) gives all three, and we need to walk them in order.
1. The mechanics of 熏習 (vāsanā / impregnation) in general
Before we reach 名言, the text sets up the mechanics of any vāsanā. The conditions are strict:
- The 能熏 (perfumer / the active consciousness that does the impregnating) and the 所熏 (perfumed / the substrate that gets impregnated) must be 俱生俱滅 — they must arise together and cease together at the same time and place. - The perfumer's continuity is shorter; once it ceases, the perfumed continues alone, undergoes a subtle change, and *that change* becomes the cause of a later arising of a similar dharma. - The metaphor the text gives is 花熏麻 — flower impregnates hemp. Flower and hemp coexist and perish together; the repeated action of the flower leaves the hemp altered, so that later the hemp produces fragrance.
The point to underline: vāsanā is not "imprinting" in a passive sense. It is an *active* modification of the substrate by an active moment of mind. The moment of mind passes; the substrate retains a transformed tendency.
2. Why only 阿賴耶識 (ālaya-vijñāna) can be the substrate
The text is unambiguous: 唯說阿賴耶識是所熏習 ("Only the ālaya-vijñāna is said to be the perfumed"). Several alternative accounts are explicitly rejected:
- 前念熏後念 (an earlier moment of mind impregnating a later moment of the same mind) — rejected as not sound. - 六識展轉相熏 (the six consciousnesses impregnating each other) — rejected. - 依剎那種類安立熏習 (impregnation based on momentary kinds) — rejected.
Why? Because the ālaya-vijñāna has the three conditions no other consciousness has:
- 相續隨轉 — it continues without break, moment to moment, life to life. - 性唯無記 — its nature is neutral/indeterminate; it is not intrinsically good or bad, so it can receive any coloration. - It is impermanent (so it can be transformed) but it can stably coexist with the active perfuming consciousnesses.
The six transforming consciousnesses (六轉識 — eye, ear, nose, tongue, body, mental) cannot serve as the substrate because their 所依 (bases), 所緣 (objects), and 作意 (intentional attention) are all distinct from moment to moment — you cannot guarantee the strict simultaneity the model requires. (In the *Saṃdhinirmocana Sūtra*《解深密經》and the *Yogācārabhūmi*《瑜伽師地論》the same argument is given.)
This is the deeper reason the ālaya-vijñāna has to be posited at all: it is the only candidate that satisfies the conditions for storing seeds.
3. What specifically gets stored: name-impregnation, not generic habit
Now we come to the distinctive part. The text — using T1597, Xuanzang's translation of Vasubandhu's commentary — gives the canonical example:
「眼名言熏習，於異熟識中為眼生因；異熟所生眼從此而生，生已還說名為眼。耳等一切名言差別應知亦爾。」 "The *eye* name-impregnation becomes the cause of the eye's arising in the matured consciousness; the eye produced by vipāka (matured result) arises from this cause, and once arisen it is still called 'eye.' All other differences of name-impregnation — ear, etc. — should be understood in the same way."
This is a small sentence doing a very large amount of work. Three things follow from it:
a. The name-impregnation is specific, not generic. There is an *eye* name-impregnation, an *ear* name-impregnation, a *nose* name-impregnation, and so on. Each is paired with the dharma it produces. The text refuses to say one undifferentiated habit-energy produces everything.
b. The names and the dharmas correspond point by point. The *difference* in names matches the *difference* in dharmas. This is the structure by which your world comes pre-sorted.
c. The eye that arises is still called "eye" because the name and the dharma continue together; the name is not a separate label tacked on after the fact. Naming and the named co-arise.
4. The two seed-types for 意識 (mano-vijñāna / discriminating consciousness)
The text then turns to the most active of the consciousnesses — the one that thinks, plans, narrates, judges, fantasizes. It is given two seed-types that explain its power:
「意識以自名言熏習為種子，一切識名言熏習為種子故，能以無邊行相分別而轉，普遍計度一切境界。」 "Because consciousness takes *its own* name-impregnation as seed and the *name-impregnation of all consciousnesses* as seed, it is able to turn with boundless aspects of discrimination and universally calculate all realms."
The two seeds are:
- 自名言熏習 (sv-nāma-vyañjana-vāsanā): seeds formed by the conceptual chatter of consciousness itself throughout beginningless samsāra. This is what makes consciousness itself arise as a *consciousness* with its characteristic discriminating function. - 一切識名言熏習 (sarva-vijñāna-nāma-vyañjana-vāsanā): seeds formed by the name-impregnations associated with the 影識 (reflected images) of boundless forms. These seeds make consciousness *appear as if* it were those forms — they are why you can be conscious *of* a mountain, *of* a person, *of* an idea.
This pair explains something easy to miss: consciousness is not a passive screen. It is a discriminating engine whose every act of awareness is fueled by stored name-impregnations. Every time you recognize anything as *anything*, name-impregnation is doing the work.
5. 多聞熏習 / 聞熏習 (multi-hearing / hearing impregnation) — a special case
Inside the same framework there is an important variant. When you hear a teaching and *correctly reflect on it* (正思), the correct reflection and consciousness coexist and perish together; consciousness gradually understands the 名句味 (the words, sentences, and meanings) heard; this impregnates the ālaya-vijñāna; when the current consciousness ceases, the next arising consciousness can grow progressively more lucid about the same material. This is why what was heard is not forgotten, why 思慧 (analytical wisdom) becomes firm and 修慧 (cultivation wisdom) becomes stable.
The crucial doctrinal nuance comes in T1598 (Asanga's commentary): 聞熏習 is not the nature of ālaya-vijñāna. It is its 對治 (antidote) — the seed of the transcendental uncontaminated mind (出世無漏心), hosted in the matured consciousness but not identical to its defiled nature.
This matters because it shows the framework is *not* fatalistic. The same store can hold contaminated name-impregnations and the antidotes to them.
6. Mutual conditioning: 異熟識 and 轉識
The verse-formula in T1598 and T1597 sums up the relation:
「言熏習所生，諸法此從彼，異熟與轉識，更互為緣生。」 "Born from name-impregnation; dharmas, these from those; the matured consciousness and the transforming consciousness mutually arise as conditions for one another."
The text illustrates this with two paired metaphors: 燈焰與燈炷 (lamp flame and lamp wick — the flame depends on the wick, the wick is shaped by the flame's burning) and 蘆束相依 (a bundle of reeds leaning on each other — neither stands alone). The ālaya-vijñāna does not sit inert. It conditions the active consciousnesses, which in turn impregnate it with name-impregnations, which in turn shape the next configuration of ālaya, moment by moment.
7. The three impregnations framework
In T1597, the full structure of ālaya is laid out with three types of impregnation-distinction (三種熏習差別):
- 名言熏習差別 (name-impregnation) — this concept - 我見熏習差別 (self-view impregnation) — seeds formed by the grasping to a self - 有支熏習差別 (existence-faculty impregnation) — seeds formed by volitional actions (業) that constitute the twelve links of dependent origination
These three are parallel, not identical. They explain three different mechanisms by which seeds in ālaya condition experience. Conflating them is one of the easiest mistakes, and we will return to it.
8. Causal continuity and internal location
Two closing pieces of the doctrine are necessary:
- Causal continuity: name-impregnation cannot arise from nothing. If a brand-new name-impregnation appeared with no prior same-type cause, an 等流果 (equal-flow fruit, i.e., a result of the same type as its cause) would be arising without a 同類因 (cause of the same type) — which is 不應道理 (illogical). There must be continuity of same-type conditioning back and back, which is part of why the ālaya-vijñāna is said to have no first beginning. - Internal location: the impregnation is in the 內 (interior), not the 外 (exterior). If external objects had no impregnation and there were also no internal seeds, then hearing and the like could not produce their results — and we would get absurdities: causes performed would lose their fruits, causes not performed would gain them. The internal location is not a metaphysical commitment; it is forced by the need to account for the regularity of how experience follows on conceptual activity.
二、生活里的走查
The classical definition in your hands — now mapped onto a Tuesday.
Scene 1 — The commute narration
It is the morning commute. You are not speaking to anyone. Yet your mind is doing something specific: it is naming, describing, judging. *This traffic is ridiculous. The driver in front is selfish. I'll be late. Late again. Why does this always happen. I'm the kind of person this always happens to.*
Let us walk this through the classical definition, step by step:
- Each of these sentences is an act of conceptual discrimination (戲論分別) — a name-and-proposition being held in mind. - Each one is a 能熏 (perfumer) acting on the ālaya-vijñāna as 所熏 (perfumed), in the strict sense: they coexist with that storehouse moment and perish together, leaving a transformation in the store. - The content of what gets stored is specific. The "traffic is selfish" thought does not store a generic "today is bad" seed; it stores a *traffic-is-an-attack-by-bad-actors* seed — with the precise coloration of grievance. The "I'm the kind of person this happens to" thought does not store a "neutral self-description" seed; it stores a *self-as-victim* seed. - That seed, as the 親因緣 (proximate cause), will later — perhaps tomorrow morning, perhaps next week — help produce another moment in which traffic is again encountered as an attack and the self is again felt as victim.
The mapping is tight. The classical definition says the differences in names correspond to the differences in dharmas produced. Here, the differences in *self-talk genres* correspond to the differences in *felt experience genres*.
The most common mis-mapping here: people think the work is in the *traffic*. It isn't. The traffic is the same traffic for everyone. The work is in the naming, which has just been deposited in ālaya as a seed, which will condition tomorrow.
Scene 2 — The phone scroll and its algorithm
You open an app. You scroll. The feed begins to know you. After a few minutes, the feed is showing you a tighter and tighter version of what you previously lingered on.
This is a useful *partial* analogy for name-impregnation, with three carefully marked limits:
| Aspect | Algorithm | Name-impregnation | |---|---|---| | Updates from your activity | Yes — every click, every dwell time | Yes — every act of naming | | Stored as a profile | Yes — vector of weights | Yes — seed (功能差別 / functional disposition), not a thing | | Conditions future output | Yes | Yes | | Operates without your awareness | Yes | Yes | | Can be "erased" by willpower | No — only by retraining | No — only by sustained retraining through antidotes (聞熏習) | | Distinguishes contaminated from uncontaminated seeds | No | Yes — that is why 聞熏習 can exist | | Has a beginningless past | No | Yes — 無始 (beginningless) |
The limit that matters most: the algorithm has a starting point — the first user profile was empty. Name-impregnation does not — the ālaya has been accumulating since beginningless samsāra. There was never a moment of "empty seed."
The most common mis-mapping here: people think the analogy proves that "the mind is just a machine, training itself." It doesn't. The analogy only shows that *some* features of conditioning-by-activity are shared. What it does not show is how antidotes (correct hearing — 聞熏習 — T1598) can produce transcendental uncontaminated seeds; nor how the *very subject* who is being trained is itself produced by the training. The algorithm has a designer and a user; the system of seeds has neither.
Scene 3 — Naming in an intimate relationship
Two people have been together for years. One of them says, almost without thinking: *You're so selfish. You always do this. You're just like your mother.*
What has happened, in the classical frame:
- Three name-impregnations just got deposited in the speaker's ālaya-vijñāna: 1. *partner-as-selfish* (a category, not a perception) 2. *partner-as-repeater* (a category about pattern) 3. *partner-as-resembling-parent* (a category about lineage) - These are specifically colored seeds — they are not "I have thoughts" seeds; they are "the person across from me is a type" seeds. - Because consciousness discriminates with 自名言熏習 (its own conceptual seeds) and 一切識名言熏習 (the seeds from reflected images), every future encounter with this partner will now be filtered through these deposits. The partner's face, gestures, words will appear *as if* confirming the category — because that is what 一切識名言熏習 does: it makes the recognized appear as if it were the type already known. - Over time, the partner begins to act in ways that *match* the name. Not because the name is true in some metaphysical sense, but because 異熟與轉識更互為緣生 — storehouse and active consciousness mutually condition each other. The seed conditions the next perception; the next perception impregnates the seed; the seed conditions the next behavior; the behavior impregnates the next seed.
The painful part of this doctrine is here: the partner has been *helped into the shape of the name* by the name itself. The name was not a passive description; it was an active impregnation that co-shaped the relationship it described.
The most common mis-mapping here: people think the question is "is the name accurate?" The classical doctrine says the more important question is "what seeds am I depositing by using this name, and what will those seeds help produce?"
Scene 4 — The first category given to a child
A small child does something clumsy — knocks over a cup. The parent says, casually: *You're so clumsy. You always do this.*
This is one of the most concrete demonstrations of the doctrine of name-impregnation. The child is at the stage of life where self-recognition is being formed. Every name the child hears about itself becomes a seed. The seed becomes a conditioning factor for the next self-recognition. Over years, the child becomes *clumsy in the way clumsy is defined* — not because of motor capacity, but because the name has shaped the field within which motor activity is recognized and produced.
Note the asymmetric structure: the child cannot yet reflect critically on the name being given (正思 / correct reflection, the function required for the antidotal path, is not yet mature). The child is in the position of pure reception of name-impregnation. This is why the *Mahāyāna-saṃgraha* and the *Saṃdhinirmocana Sūtra* are so careful about *what kind of name is being deposited by whom.*
The most common mis-mapping here: people think the parent is just "describing." The classical doctrine says the parent is impregnating — that is, doing an active causal work on the substrate of the child's ālaya-vijñāna that will condition years of felt experience.
Scene 5 — Hearing a teaching
You sit down and read a piece of writing that re-names something you had been naming badly. Perhaps for years you have thought of your own anger as "a sign of strength" or your own avoidance as "self-protection." The teaching offers: anger as the fire of grasping; avoidance as a contraction against the impermanence of things.
What the classical text would call 多聞熏習 (multi-hearing impregnation) begins to operate:
- The 正思 (correct reflection) and your consciousness coexist and perish together. - The new 名句味 (words, sentences, meanings) gradually become clear in your mind. - This impregnates your ālaya-vijñāna with a *contaminated-but-antidotal* seed — T1598 calls it the 對治種子 (antidotal seed), capable of producing 出世無漏心 (transcendental uncontaminated mind). - The first moment of consciousness carrying the new naming may be weak. The next moment will be slightly stronger. The next stronger still. This is the 次第增勝 (progressively increasing) structure that the *Mahāyāna-saṃgraha* describes.
This is the hopeful side of the same doctrine. The same store that has been filled with "I am the kind of person this happens to" seeds can also receive seeds of "anger is the fire of grasping" — and those seeds, because they are antidotal, work *against* the defiled name-impregnations rather than merely adding to them.
三、当代人为什么需要它
Why this obscure technical concept should matter to you now.
1. Because the rate of naming has accelerated beyond what any previous period has known
A person in a pre-literate village did perhaps several hundred acts of conceptualization per day, many of them ritualized and shared. A person now does tens of thousands — through feeds, captions, internal commentary on the feed, the feed's responses, the messaging, the messaging's emotional narration, the narration's narration. Each one is a potential impregnation.
You are not *doing something new* by being on a feed. You are *doing vastly more* of what has always been happening. And the doctrine of name-impregnation says that *the rate of doing* matters causally — not in a vague "practice makes perfect" sense, but in the strict sense that each act leaves a transformed substrate.
This is one of the cleanest pieces of evidence for taking the concept seriously: when the rate of any causal activity goes up by orders of magnitude, the cumulative causal effect on the substrate goes up by orders of magnitude.
2. Because the *invisibility* of conceptual shaping is the very problem
The classical doctrine is explicit: name-impregnation works below the threshold of noticing. The *Mahāyāna-saṃgraha* says consciousness discriminates "with boundless aspects" and "universally calculates all realms" — that is, the naming activity is happening at every moment of conscious life, and most of it is not surfaced as "I am currently naming."
A contemporary person often believes their preferences are *discovered* ("I just don't like mornings"), their reactions are *natural* ("I'm just an anxious person"), their judgments are *perceptual* ("he really is selfish"). The doctrine of name-impregnation offers a precise way to read this: what you are calling discovery, nature, and perception is largely the activity of stored name-impregnations being read back as if they were the way things are.
To see this is not to be cynical. It is to be precise.
3. Because the antidote structure is built into the doctrine, not tacked on
The same classical text that describes name-impregnation describes 聞熏習 as its antidote. The two are not separate teachings; they are paired. If you take the doctrine seriously, you must take both halves seriously — which means: yes, your inner narration is shaping you; *and* you can change that narration by exposing yourself, repeatedly and reflectively, to a teaching that re-names accurately.
The contemporary resonance is sharp. The classical text is saying: every act of naming is a vote for the kind of world that will appear to you tomorrow. The antidote is not "stop voting" (you cannot stop — the framework says conceptual activity is beginningless and ongoing). The antidote is *changing the ballots*. And the ballot-box that matters is not the speech you give to others — it is the 正思 (correct reflection) you bring to what you hear.
4. Because liberation is reframed as seed-work
This is the deepest contemporary relevance. In the framework of the *Mahāyāna-saṃgraha*, the path is not primarily about deleting bad thoughts or suppressing bad behavior; it is about transforming the seed-configuration of the storehouse. The active consciousness cannot be reformed in isolation; it is sustained moment by moment by ālaya's seeds. The seeds cannot be reformed in isolation; they are continually being replenished by active consciousness.
This is a *systems* view, not a willpower view. It explains why willpower alone fails (you cannot impregnate ālaya with a single heroic act against the weight of countless prior impregnations), and it also explains why gradual sustained practice works (each new antidotal seed, because it is the same-type cause of further antidotal moments, conditions the next, in a positive feedback loop).
5. Because it gives you a usable map of how language is doing something *to you*
There is a famous modern line of thought — sometimes called the Sapir-Whorf hypothesis in linguistics, sometimes implicit in discourse analysis — that language shapes perception. The classical doctrine is more precise than these: it specifies that it is not language in the abstract, but *acts of naming and conceptualization*, happening in the specific technical form of 能熏/所熏 (perfumer/perfumed) with 俱生俱滅 (co-arising and co-ceasing), that deposit seeds in a specific substrate (ālaya-vijñāna), which then condition specific future experiences.
Modern linguistics can show correlations between language and perception; it cannot yet say where the *causal storage* is happening. The classical doctrine makes a specific claim about *where*: in the storehouse consciousness, in the form of 功能差別 (functional dispositions), not as little physical tokens.
You do not have to accept the metaphysics to use the model. The model itself is a precise tool: every act of naming is causally downstream and causally upstream of your felt experience. Use the upstream handle deliberately and the downstream changes.
四、常见误读与澄清
Misreading 1 — "Name-impregnation is just verbal habit, like being told 'don't say that' as a child"
The classical concept is much narrower in one direction and much wider in another.
- Wider: name-impregnation includes not just words but the entire apparatus of 名 (concept), 句 (sentence/proposition), 味 (meaning). It also includes 影識 (reflected image) — the silent recognition of a face, a place, a tone as *of a certain type*. The text's two seed-types for consciousness — 自名言熏習 and 一切識名言熏習 — together cover both the verbal and the imagistic. So "verbal habit" under-describes the scope. - Narrower: name-impregnation is *not* the same as volitional action (which is 有支熏習 / existence-faculty impregnation) and *not* the same as self-grasping (which is 我見熏習 / self-view impregnation). You can change your words without changing the conceptual seeds, and you can change the conceptual seeds while outwardly producing the same words.
Misreading 2 — "It's the same as karma"
It is not. Karma, in the technical sense of 有支熏習, refers to the volitional actions (intention + action) that constitute the 十二有支 (twelve links of dependent origination). Name-impregnation is conceptually different: it is the conceptual-symbolic apparatus that shapes *how* those actions are framed and *what* gets recognized as a possible action at all. A being with no name-impregnation could not even form the intention to act in the way they do — the very categories of "self" and "other" and "action" and "goal" are name-impregnations.
So name-impregnation is *logically prior* to volitional karmic seeds in one sense, while coexisting with them in another. The *Mahāyāna-saṃgraha* lists all three together (T1597) precisely because they are not the same.
Misreading 3 — "Seeds are little things stored in a 'mind-warehouse'"
The text explicitly refuses this picture, especially in T1598: the seeds are not "另有實物住在識中" ("separate real objects dwelling in the consciousness"). They are 功能差別 (functional dispositions) of the ālaya-vijñāna itself — the consciousness's *capacity* to produce certain kinds of dharmas. To picture them as little colored balls in a jar is to import a substance-metaphysics the doctrine explicitly does not endorse.
The accurate picture: ālaya-vijñāna is like a field whose properties have been modified by everything that has ever grown in it; those modifications are not things added to the field, they *are* the field's present disposition.
Misreading 4 — "You can just stop naming and the seeds will fade"
This is the willpower misreading. The classical framework is explicit that the impregnating activity is beginningless (無始) and ongoing. You cannot stop naming any more than you can stop your heart. The path is not "stop"; the path is 改名作 (changing what you name and how you name).
This is why the antidote — 聞熏習 — is not a one-time event but a sustained co-arising of correct reflection (正思) with consciousness, gradually impregnating the antidotal seeds, which then condition the next moment's perception, which impregnates the next seed, in a positive feedback loop.
Misreading 5 — "If it's all seeds, there's no freedom"
The opposite is closer to the text's implication. The storehouse does not act on its own; it acts 更互為緣 (mutually with) the active consciousnesses. The active consciousnesses are not determined; they take their seeds as conditions but their arising is fresh each moment. And the antidote — correct hearing, correct reflection — is precisely a *new kind of condition* that enters the mutual loop and gradually re-shapes it.
The text's image of 燈焰與燈炷 (lamp flame and lamp wick) is doing work here: the wick conditions the flame (no flame without wick), but the flame also shapes the wick (burns it down, requires trimming). Neither side is unconditioned; neither side is fully determined. The mutual conditioning leaves room for the gradual transformation that the bodhisattva path describes.
Misreading 6 — "This is just Buddhist psychology; it doesn't really say anything metaphysical"
Read T1597 carefully: 「阿賴耶識…虛妄分別種子，及一切虛妄分別生起因緣」("the ālaya-vijñāna [is] the seed of false discrimination, and the cause-condition by which all false discriminations arise"). This is a metaphysical claim — that there is a substrate whose function is to be impregnated by conceptual activity, and that this substrate has no first beginning. You can use the model pragmatically without affirming the metaphysics, but you should be honest that the model is offered within a metaphysics, not as a neutral description.
Misreading 7 — "Modern neuroscience has proven this"
It has not. What neuroscience has shown is that neurons that fire together wire together (Hebbian learning) and that long-term potentiation (LTP) is a real mechanism for memory consolidation. These are real physical mechanisms for how repeated activity modifies a substrate.
But the classical doctrine makes at least four claims neuroscience has not addressed:
- That the substrate is *beginningless* — neuroscience sees LTP developing; it does not see an LTP without a first LTP. - That there is a strict 俱生俱滅 (co-arising/co-ceasing) requirement between perfumer and perfumed — neuroscience does not require this for memory formation. - That there are *contaminated* and *uncontaminated* seeds, the latter being antidotes to the former (聞熏習, T1598) — this moral/transformative dimension is not part of the LTP framework. - That the very *subject* doing the conceptualizing is itself produced by the conceptualizing — neuroscience studies brains, which are physical objects; it does not claim brains are produced by their own firing patterns in the way ālaya is claimed to be produced by its own impregnations.
The analogies are illuminating; they are not proof.
收束
A small working summary to take with you:
Every act of naming — spoken, thought, or silently recognized — is, in the *Mahāyāna-saṃgraha*'s terms, an active moment of consciousness leaving a specific, color-shaped transformation in the storehouse. That transformation is the seed that, with the cooperation of the active consciousnesses and the matured consciousness, helps produce the next moment's experience of the world. The names and the dharmas co-arise; the differences in names correspond to the differences in dharmas.
The path is not to stop this — it cannot be stopped — but to change what is being named and how, sustained over time, until the antidotal seeds are strong enough to transform the configuration. The classical text does not promise quick results; it promises a precise mechanism, a paired antidote, and the long patience of 次第增勝 (progressively increasing).
In a period of life where the rate of conceptual activity has multiplied, and where the apparatus of naming is no longer just your own but the entire feeding architecture of the attention economy, this old doctrine is not less relevant. It is more.
And the most personal version of it is the simplest: what are you naming, right now, in the inner monologue that nobody else hears — and what seeds is that naming depositing in the store you will inhabit tomorrow?