A note on where this comes from: this concept is set out with characteristic precision in 《阿毘达磨俱舍论》(*Abhidharmakośa-bhāṣya*, T1558), the great compendium of the Sarvāstivāda-Vaibhāṣika school (毘昙部). The tradition's whole apparatus of analysis grew up around one stubborn question, and 等起 is one of the answers. If you read it carefully, you can hear a particular kind of mind at work — the mind that wants to understand why good people do bad things, why bad people sometimes do good, and why what we call "an action" is never quite as simple as a body moving through space.
1. What this concept is actually about
The puzzle behind the framework
A body moves. A voice speaks. From the outside, that's an "action." But somewhere inside, something stirred first. The classical formulation is blunt about this: a bodily or verbal act does not just *happen*. It has to be set in motion by a mind, and accompanied by a mind — otherwise it would be no different from a corpse twitching. The word for this launching-and-accompanying mechanism is 等起 (samutthāna), literally "rising together with" or "jointly arising."
The reason the term has to be technical is that the early Buddhist analysis was trying to solve a problem that every moral tradition faces: What gives an act its moral color (善、不善、无记 — wholesome, unwholesome, or neutral)? Is it the act itself? The agent? The intent? The consequences? The Abhidharma answer is "all of the above — but in distinct, named ways."
The four-cause framework: where 等起 sits
The *Kośa* organizes this by naming four causes through which any dharma (any phenomenon) comes to have a moral quality:
- 胜义 (*paramārtha*, ultimate/definitive) — the inherent nature of a phenomenon in itself. - 自性 (*svabhāva*, own-nature) — the dharma's own defining characteristic. - 相应 (*saṃprayoga*, association) — the way mental factors accompany each other within a single mind-moment. - 等起 (*samutthāna*, joint arising / impulse) — the way a prior or simultaneous mind launches and sustains a bodily or verbal act and gives that act its moral color.
The classical formulation in the *Kośa* is worth quoting directly:
诸法可由胜义、自性、相应、等起四因而成善等性。
Translated: "All dharmas can become wholesome (or otherwise qualified) by way of four causes: ultimate, own-nature, association, and impulse."
等起 is not the whole answer. It is the answer to *one specific* question: how does a piece of behavior — a wave of the hand, a sentence spoken — get its moral color from the mind? The other three causes deal with other aspects (the act's own nature, the moment-by-moment co-arising of mental factors, the ultimate ground). 等起 is the one that handles the mind-to-body-to-world interface.
The two layers of 等起: 因等起 and 刹那等起
Here the analysis does something unusual and important. It splits 等起 into two distinct moments, which together account for everything a mind does when it produces an action:
因等起 (*kāraṇa-saṃutthāna*, causal impulse) — the prior mind that initiates the act. You decide, before the act, to do it. This is the moment of resolve, of "I will now do X." The text calls this 转 (*pravartana*, "turning toward / initiating"). It is the mind that *turns toward* the action, points itself at it, launches it.
刹那等起 (*kṣaṇa-saṃutthāna*, momentary impulse) — the simultaneous mind that accompanies the act while it is happening. The text calls this 随转 (*anuvartana*, "going along with / accompanying"). It is the mind that is *actually present* as the act unfolds.
The *Kośa* gives the definitions in plain terms:
因等起者，谓将欲造作身语业时，先行作为原因而能引发该业；刹那等起者，谓正造业时与业不相离。
Translated: "Causal impulse is the prior mind that, when one is about to do a bodily or verbal act, stands as the cause and brings that act into being. Momentary impulse is the mind that, in the very moment of acting, is inseparable from the act."
This is why the concept needs two names. "Deciding to apologize" and "the apology, while it is being spoken, with whatever tone of voice and degree of presence" are not the same mental event. The classical analysis insists you keep them separate, because the moral story is told across both.
Why both layers are necessary — the dead-body analogy
The *Kośa* offers a strikingly vivid reason why 刹那等起 cannot be omitted. It says:
虽已有先前的因心引发行为，若造业当下没有随转心，行为便应如死人之业而不能成立。
Translated: "Even if a prior mind has initiated an action, if at the moment of acting there is no accompanying mind, the action should be no different from an action of a corpse — it could not actually be constituted as an action."
This is more than a flourish. It says something philosophically important: **a body doing something with no live mind present is not, properly speaking, an *act* in the moral sense.** The corpse analogy is doing real work here. It is the same kind of move that, in contemporary moral philosophy, distinguishes an "act" from a mere "event" — only a being with present awareness can be said to *do* something rather than merely *have something happen to it*. 等起's insistence on 刹那等起 protects the live, moment-by-moment presence of mind as a non-negotiable part of what makes something a real action.
The fourfold map of which minds can do what
Not every mind can be both the initiator and the accompanier. The *Kośa* lays out a careful map of which kinds of consciousness can fill which role, in four cases:
有转非随转，有随转非转，亦转亦随转，非转非随转。
Translated: "Some minds are initiators but not accompaniers; some are accompaniers but not initiators; some are both; some are neither."
What fills each slot is technical but very concrete:
(a) Initiator but not accompanier — 转非随转: 见所断识. These are minds afflicted by *views that need to be abandoned through seeing* — that is, minds saturated with wrong views about self, persons, etc. Such a mind can launch an action ("I, a self, will now do X"), but it cannot simultaneously accompany the act with proper attention. It is the mind of an act of violence done from a hardened ideology, perhaps: the launching is there, but the live moment-by-moment registration is distorted by the view.
(b) Accompanier but not initiator — 随转非转: the five sense consciousnesses (眼等五识). Your seeing, hearing, smelling, tasting, touching. These are pure sensory contacts with objects. They have no conceptual discrimination and no built-in intention to launch anything. They simply register. But every time an act is performed, the five senses are *accompanying* it — you see what you are doing, hear yourself speak, feel the resistance of the surface. They are never the initiator.
(c) Both initiator and accompanier — 亦转亦随转: 修所断的善、不善、无记三性意识. These are minds of the three moral qualities (wholesome, unwholesome, neutral) that are *cultivable*, capable of being refined or abandoned through practice. They have both discrimination and outward engagement, so they can both launch and accompany acts.
(d) Neither initiator nor accompanier — 非转非随转: 无漏心 and 异熟生心. *Anāsrava* minds (undefiled, fully purified) and *vipāka-ja* minds (the mind that arises as the mere ripening of past karma, without fresh volition) do not function as initiators or accompaniers of ethical action in this technical sense. The first because they operate beyond this whole mechanism; the second because they are too passive, too merely-receptive.
This fourfold map is not abstract. It tells you something about why you can be in a situation where your senses are tracking an act you feel you didn't really choose. Your senses are the 随转; the choice was made earlier by a different kind of mind.
The key surprise: 转 and 随转 can have different ethical colors
Here is where the concept cuts against intuition. You might assume that the initiating mind and the accompanying mind would naturally be the same — you decide to be kind, and you are kind while doing it. The *Kośa* insists that this need not be the case. The same passage that lays out the fourfold map also says:
前一转心若为善，后一随转识仍可通于善、不善、无记；转心为不善或无记时，也可有相应差别。
Translated: "If the prior initiating mind is wholesome, the subsequent accompanying consciousness may still be wholesome, unwholesome, or neutral. The same variety holds when the initiating mind is unwholesome or neutral."
You decide to help someone (wholesome initiator). You sit down to actually help, and your mind drifts — you are bored, distracted, mildly irritated by them (unwholesome or neutral accompanier). The act's moral color is determined by the initiating mind; the moment-by-moment texture is determined by the accompanying mind. They are two different facts.
The *Kośa* notes a special case: the Buddha. Even for him, the text says, the 转 and 随转 are mostly the same, but "少分不同" — partially different. It also records a *disputed* position from "some other school" that the Buddha's mind is permanently in samādhi and never has neutral moments, hence his 转 and 随转 are entirely the same. This is a sectarian disagreement the *Kośa* reports but does not resolve.
How the actual ethical color of an act is fixed
The last piece of the technical puzzle: when you actually see an action and want to know whether it is wholesome, unwholesome, or neutral, which mind do you look at?
The answer the *Kośa* gives is the 修所断转心 — the *cultivable* initiating mind. Why specifically this and not the 见所断转心 (the view-corrupted one)? Because the *Kośa* argues that the 见所断转心 and the actual bodily/verbal act have a 修所断 mind *intervening between them* — a slice of ordinary, non-view-corrupted consciousness sits in the middle. So the act's color is fixed by what is closest to it: the cultivable, ordinary initiating mind, not the more ideological, view-corrupted one further upstream.
This is a subtle but important point. It means that your most deep-seated worldview does not directly paint your actions in its colors — there is always a layer of more ordinary, more practical, more corrigible mind in between, which is where the moral attribution actually settles. This is a structural argument for why people with deep wrong views can still act with ordinary decency, and why people with correct views can still act badly — the deciding factor is the *practical* initiating mind that is close to the act, not the *view* behind it.
2. Life walkthrough
The classical categories are dry-looking on the page. Let me try to walk one full contemporary situation through them, slowly, and then add two shorter ones.
A long walk: the moment you snapped at your child
You have had a long week. The project at work is in a frantic phase. You slept poorly. There is a low-grade hum of resentment toward your partner for something unsaid. You are driving home. None of this is in words; it is a heaviness in the body, an impatience with the traffic, a low-level scan for what will go wrong next.
This heaviness is the 因等起 already taking shape. It is not yet any *decision* to do anything. It is the prior condition that will become capable of launching an act. In classical terms, it is the kind of mental field that 修所断意识 can operate in.
You walk in the door. Your six-year-old runs up to greet you, full of a story about something that happened at kindergarten. You hear about two minutes of it. Then they reach for the glass of milk on the table, you see it tip, and the milk spreads across the kitchen floor.
What happens in the next second is, technically, four things happening in quick succession.
Step 1 — the five senses register. The eyes see the white milk spreading. The ears hear the small clink of the glass and the child's small surprised "oh." The hand feels the weight of your bag. These are the five sense consciousnesses — 眼识、耳识、身识 — and they are purely 随转 (accompanying), never 转 (initiating). They cannot decide anything. They simply register.
This matters because, in the moment of crisis, a lot of moral weight gets laid on "what you saw" or "what you heard." But the *Kośa* is explicit: the senses don't initiate. They can only accompany. If you later say "I just saw red," that is a figure of speech — what actually saw the milk was an innocent sensory contact.
Step 2 — the interpreting mind turns toward. Almost instantly, the *interpretive* mind — 意识 — labels what just happened. "This is the third spill this week." "He never listens." "I can't even come home without this happening." This is the mind that has both discrimination and outward engagement — 修所断意识. It is now capable of being both 转 and 随转.
In this moment it is doing both. It is interpreting (which is part of accompanying what is happening) and it is turning toward the action that is about to happen — it is the moment the mind says, internally, "I'm going to say something now." That inner "I'm going to" is the 转 in its most nascent form.
Step 3 — the initiating mind crystallizes. A moment later, the decision hardens. "I need to teach him a lesson about carelessness." This is the 转心 at full strength. In the *Kośa*'s framework, it is a 修所断转心 — a cultivable initiating mind, not yet a view-corrupted one, but unmistakably unwholesome in color. It is the mind that will paint the next act.
Step 4 — the act unfolds, accompanied by a different mind. Your voice comes out sharper than you planned. The actual words are: "Look what you've done *again*! How many times have I told you? You are so careless, you never think!" You feel the words leaving your mouth. You see the child's face fall.
Here is the 等起 surprise: the *mind that accompanies this act* — the 随转 — is not quite the same as the *mind that initiated it*. Your initiating mind was full of pedagogical conviction: "I'm teaching him a lesson." The accompanying mind, in the actual moment of speaking, is mostly exhaustion and a kind of animal irritation. The 随转 here is closer to neutral or mildly unwholesome — it is the tiredness acting through your voice, the impatience coloring the tone, the disappointment hitting the child's face.
If a contemplative observer were watching you at this moment, they would see two different mental events stacked: a *committed* initiating mind a few seconds before, and a *reactive, tired* accompanying mind during the act. Both are real. Neither is the whole story.
Step 5 — the moral color settles. The act of speaking — what the *Kośa* calls the 表业 (expressive/visible act) — has its moral color fixed by the 修所断转心, the cultivable initiating mind. The act is unwholesome. Not because of the words as such, not because of the spilling, not because of the milk. Because of the mind that launched it.
But the *texture* of the act — how it lands on the child, what it carries in its tone — is shaped by the 随转 mind, the exhaustion. This is why, when you replay the scene later, you might be uncertain whether to blame yourself for "what you said" or for "how tired you were." Both. The classical analysis says: both are real, and they sit in different layers.
Step 6 — the after-image. Ten minutes later, you see the child still not smiling. The five senses register that. A new round of 意识 begins. A new possible 转 starts forming: "I should apologize" or "I should justify" or "I should pretend nothing happened." 等起 is never a single act; it is a chain of initiations and accompaniments, layer after layer, throughout the day.
A short walk: posting in anger online
You see an article you disagree with. You read it. The 五识 (eye-consciousness, etc.) register the text — purely accompanying.
A comment thread offends you. The interpretive 意识 labels it: "these people are idiots." This is a 修所断意识 acting as both 转 and 随转.
You decide to compose a reply. The 转心 is indignation. The 随转, as you type, drifts — between sentences you check another tab, you think about dinner, you wonder if this is even worth it. The 随转's quality is different from the 转's: less hot, more distracted.
You press send. The 表业 (the public act of speaking — and in the *Kośa*'s framework, "speaking" includes any way the body communicates) gets its moral color from the 转心 (indignation), but the words themselves carry the 随转's distractedness — they come out more rambling, less sharp, than the indignation alone would have produced.
The replies start arriving while you are brushing your teeth. New initiations. New accompaniments. The chain goes on.
The thing most people get wrong here: they think the question is "should I have sent that?" The 等起 analysis says the more useful question is two questions stacked. *What was the mind that launched it?* And *what was the mind that actually accompanied the typing?* Often these are not the same mind at all, and the difference is the difference between feeling righteous and feeling embarrassed at the message later.
A short walk: a moment of uncharacteristic kindness
You are in line at a coffee shop. The person ahead of you is fumbling for change, short by a small amount. You say "I'll cover it" before you have really decided.
What just happened? In the classical terms, the 转 — the prior launching mind — was probably quite minimal: a kind of reflexive, not-very-thoughtful goodness, perhaps closer to neutral than to deeply wholesome. There was no time for a "decision." But the 随转 — the mind accompanying your actual speaking and reaching for your wallet — is genuinely warm. You see them smile. You feel the small warmth of having done a small thing without ceremony.
The act's moral color is fixed by the 转 (the cultivable initiating mind, near-neutral). The *felt* moral texture is carried by the 随转 (the warm accompanying mind). The two do not match. Most people, later, remember only the warmth. The classical analysis insists: the warmth was the 随转, and the launching was something more ordinary. Both belong to the act.
3. Why contemporary people need this
We live in an age of dissociated action
The classical account was written for monks who lived in a small world of deliberate, face-to-face actions. You bow, you speak, you eat, you walk, you meditate. Each act is brief, embodied, witnessed.
Our world is the opposite. We type and send. We click and publish. We schedule and forget. We swipe and the algorithm decides. The moral life of a contemporary person is dominated by acts in which the launching mind (the moment of deciding to post, deciding to buy, deciding to archive, deciding to deploy) is *radically* separated in space and time from the accompanying mind (the moment the words are actually read, the purchase actually arrives, the algorithm actually fires).
The four-cause framework, and especially the 因/刹那 split, is unexpectedly useful here. It says: every act you do has two mental moments, and they can have different moral colors. When you send an email at 11 p.m. in frustration and it is read at 9 a.m. by someone in a completely different mood, the moral color of *the email* is fixed by the launching mind — yours, at 11 p.m., tired. The way it lands is shaped by the accompanying mind of *the reader*, at 9 a.m., well-rested. These two minds never meet. 等起's insistence that initiating and accompanying are real and separate is, in a way, a description of the moral structure of asynchronous digital life.
It clarifies the intention-versus-execution problem
Modern moral psychology has a robust literature on the gap between *intent* and *execution* — the famous "intention-behavior gap," in which people fail to follow through on their stated intentions because at the moment of action, the surrounding conditions have changed. People intend to be patient and lose their temper. People intend to eat healthily and eat the donut.
The classical analysis maps onto this, but adds a layer most modern accounts miss: it distinguishes the mind that launches (intention) from the mind that accompanies the execution (the live moment of presence), and says that the moral color of the act is set by the launching mind — even though the *texture* of what actually happens is set by the accompanying mind.
This is non-trivial. It means: you are morally responsible for what you launched, even when what you launched came out differently from what you imagined during execution. The trans夜 email is yours, even if you would not have written it in the morning. The scolding is yours, even if you were too tired to have meant exactly what you said. The classical analysis takes responsibility seriously — it does not let you off the hook by saying "I was tired" — but it also acknowledges that the texture of what happened is part of the moral story.
It supports the practice of presence
Most of what passes for "mindfulness" in contemporary culture emphasizes *being present in the moment of action*. The 刹那等起 / 随转 distinction is, in a sense, the abhidharmic justification for why that emphasis is technically correct. If the moment of execution truly requires a living, accompanying mind, then the quality of that accompanying mind is a real part of the moral quality of the act. Being present is not an aesthetic preference; it is a moral structure. When you are half-present, you are letting the execution of your acts be carried by an under-resourced mind.
The classical analysis here is more precise than the popular mindfulness version. It does not say "be present and you will be good." It says: when you act, the mind that accompanies the act is a *real* causal factor. You can launch an act with a great initiating mind and then accompany it poorly. You can launch an act with a confused initiating mind and accompany it with genuine presence and warmth. Both happen. The implication is that cultivation of the *accompanying* mind is its own discipline, not just an offshoot of cultivation of *intention*.
It offers a more honest framework for moral responsibility than either blame or excuse
Western moral and legal traditions tend to oscillate between two extremes: the intentionalist view (you're responsible only for what you intended) and the strict-liability view (you're responsible for the actual outcomes regardless of intent). Both are inadequate to actual human life.
The classical analysis, by distinguishing 转 from 随转 and treating them as two distinct moral moments, offers a third option: you are responsible for what you launched (the 转), and the execution (the 随转) is a separate layer of reality that you can cultivate independently. This makes moral improvement more granular. You don't have to overhaul your whole intention-structure to be a better person in the moment of execution; you can work on the accompanying mind alone. Conversely, you can't get off the hook just by having had a good accompanying mind if the launching mind was callous.
In AI-related contexts — and this is just an analogy, not an equivalence — there is a similar emerging distinction between policy (the launching decision, often made by humans, often far in advance) and execution (what the system actually does in the moment, often determined by other layers). The classical 四因 framework, with 等起 at one corner of it, has the right shape for thinking about this kind of layered responsibility, even though the analogy has obvious limits: an AI system does not have a 刹那等起 mind in the classical sense, and the analogy should not be pressed.
It is a tool for self-understanding, not for self-judgment
There is a final, gentle use of this concept. When you are trying to understand why you did something you regret, the classical analysis invites you to ask not one but a small stack of questions:
- *What was the heaviness, the prior condition, that was already forming before the act?* (因等起) - *What was the actual moment of launching? What did I tell myself I was doing?* (转心) - *What was my mind actually like as the act unfolded — tired, distracted, hot, cold, somewhere else?* (随转) - *Is the color of the act fixed by the launching mind or the accompanying mind?* (the 修所断转心 rule)
These four questions will, more often than not, return different answers. That is not a failure of self-knowledge. It is, the analysis insists, the actual structure of any moral event. Most of our suffering about past acts comes from assuming it should have been simpler than it was.
4. Common misreadings and clarifications
Misreading 1: "等起 just means intention."
This is the most common flattening. In contemporary usage, "intention" usually means a vague prior state of wanting to do something. The classical term covers that, but it is much more technical. It includes (a) the *prior* initiating mind, (b) the *simultaneous* accompanying mind, (c) a structured map of *which* kinds of consciousness can be initiators and which cannot, and (d) the technical rule that the moral color is fixed by the cultivable initiating mind specifically. Calling all of this "intention" empties it of its analytical content.
Misreading 2: "This says action doesn't matter — only the mind does."
Almost the opposite is true. The whole framework of 等起 exists to explain how action *gets* its moral color. The conclusion is that action's moral color depends on the mind that launches and accompanies it, not that the action is dispensable. The body-and-voice are not optional. They are *what gets its color from the mind.* Reading this as "Buddhism says only intentions count" is a misreading that would have been rejected in the *Kośa* itself, which spends most of its length analyzing bodily and verbal acts.
Misreading 3: "转 and 随转 must be the same moral color."
Many people assume that if you started out kind, you were kind throughout. The classical text explicitly denies this. The initiating mind and the accompanying mind can differ in moral color, and in some cases — like the Buddha, on the disputed account — they can partially differ even in the most purified beings. This is not a loophole. It is a structural description of how acts actually unfold.
Misreading 4: "This is a moral ranking of people."
The fourfold map (转非随转 / 随转非转 / etc.) is a classification of *kinds of consciousness*, not of people. Your five senses are always 随转非转; that does not make you a worse person than someone whose consciousness is mostly 见所断. The map is technical. Reading it as a moral hierarchy of persons is a category mistake.
Misreading 5: "This means I should always be monitoring my initiating mind."
Some readers, having encountered this framework, try to "catch" their initiating mind at every moment and find it already morally suspect. This is a misunderstanding of the practice. The framework is analytical — it gives you the structure to understand what happened after the fact, and to recognize the structure in the moment. It is not a permanent self-surveillance device. Used as self-surveillance, it produces a different kind of suffering rather than freedom. The text's deeper interest is in *cultivating* the 修所断 mind — the ordinary, corrigible mind that is between your deepest views and your acts — so that what you launch, and how you accompany it, can gradually become more coherent and more kind.
Misreading 6: "The 刹那等起 / 随转 is just 'being in the moment' or 'mindfulness'."
Contemporary mindfulness traditions have their own rich analyses of present-moment attention. The classical 刹那等起 is *related* to that, but it is not the same. 刹那等起 is specifically the *mind inseparable from the bodily or verbal act in the moment of its performance.* It is narrower than "mindfulness of everything" — it is the live mental accompaniment to a specific act. And it is wider than "noticing your breath" — it is what is actually happening in the mind while you are speaking, working, deciding. The mapping is suggestive but not identical.
Misreading 7: "AI will eventually have 等起, so we will need this framework for them."
A contemporary temptation. It is worth saying plainly: 等起, in the classical sense, requires a mind that can be both the *prior cause* of an act and the *live accompanier* of its unfolding. Whether or not current or future AI systems can be described in those terms is an open philosophical question, and the *Kośa*'s framework was developed for a specific analysis of human (and, on some interpretations, non-human) consciousness. Drawing analogies can be productive; drawing conclusions about whether "the algorithm has a 刹那等起" is to confuse a category.
A closing thought
等起 is, at its heart, a refusal to let moral life be simple. It refuses the simple story "I did a bad thing because I am bad." It refuses the opposite story "the act was just an event, my mind is irrelevant." It insists that between you and your acts there are two distinct mental moments — one launching, one accompanying — and that the moral color of what you do is fixed by the launching, while the texture of what you do is shaped by the accompanying. It also insists that this whole picture is itself subject to cultivation: the cultivable, ordinary mind between your deepest views and your acts is where the moral weight actually settles, and that mind can be trained.
For someone living in a contemporary world of delayed, asynchronous, screen-mediated, algorithm-shaped action, this is a remarkably fitting lens. The world has made it very easy to launch acts whose execution you never quite witness. The classical framework, written for a different kind of life, still has the right shape for asking: *what was the mind that sent this? And what was the mind, or minds, that accompanied it through?*
These two questions are not optional. They are the structure of any act at all.