What This Concept Is Actually Saying
When the early discourses speak of "不善行" (akuśala-caryā — unwholesome conduct), they are not drafting a moral code in the sense we usually hear the word "moral." They are doing something more diagnostic. They are pointing to a particular *quality* of action — a quality that can be present in anything you do with your body, anything you utter, and any internal stirring in your mind — and they are saying: when this quality is present and you have done it on purpose, suffering follows. Not might follow. Not could follow if some deity notices. Follows, the way heat follows when you put your hand near fire.
The classical definition has three elements that must stay together. The *Madhyama-āgama* (《中阿含经》) puts it this way: a deed must be 故作 (deliberately done, *krta*) to count as 业 (karma — volitional action with consequences). Unintentional harm — stumbling into someone, speaking before you think, an instinctive flinch — is not the same kind of thing, and the texts explicitly note that "不故作业" need not ripen. Then the deed must be 不善 (akuśala — unwholesome, harmful, productive of suffering), as opposed to 善 (kuśala — wholesome, beneficial, productive of happiness). And third, because it is both *deliberate* and *unwholesome*, it 招感苦果 (ripens into bitter fruit) — in this life, or in a future one, or in both. The texts stress the inescapability of this third feature with an image so harsh it sounds like rhetoric until you sit with it: *"宁吞热铁丸，不可犯戒而受世间信施"* — better to swallow a ball of red-hot iron than to break the precepts while accepting the faithful offerings of the world. The point is not theatrics; it is that the burning is internal, in the ripening, and cannot be outrun by anyone, anywhere.
Where does this concept sit in the larger map of the path to liberation (解脱道)? It sits at the hinge between cause and consequence. 善行 (kuśala-caryā — wholesome conduct) and 不善行 are the two main streams by which ordinary beings continue to generate karma that keeps them cycling in saṃsāra. Understanding unwholesome conduct — really understanding it, not just labeling it — is the prerequisite for the training in ethical conduct (尸罗, *śīla*) that grounds all meditation (定, *samādhi*) and all liberating wisdom (慧, *prajñā*). You cannot see clearly if your own actions are still pouring mud into the stream. So this is not a peripheral concept. It is the foundation stone of practice, and getting it wrong at the foundation means the whole structure drifts.
The Two Main Classifications: Ten Categories, and Three Doors
The classical sources give the same content in two slightly different cuts. Both are worth keeping straight.
**The Ten Unwholesome Actions (十恶 / daśākuśala), as listed in the *Ekottara-āgama* (《增壹阿含经》)**:
1. 杀生 (prāṇātipāta) — taking life 2. 不与取 (adattādāna) — taking what is not given 3. 婬泆 / 邪婬 (kāmesu micchācāra) — sexual misconduct 4. 妄语 (musāvāda) — false speech 5. 绮语 (saṃbhinnapralāpa) — frivolous, empty talk 6. 两舌 (paisunya) — divisive speech 7. 鬪乱彼此 — inciting discord between people 8. 贪他 (abhidhyā) — covetousness, wanting what belongs to others 9. 起恚 (vyāpāda) — ill-will, wrath 10. 邪见 (micchādiṭṭhi) — wrong view
**The Three Doors (三业), as laid out in the *Madhyama-āgama* (《中阿含经》)**, sorted by body, speech, and mind:
- 身业 (body deeds, three): 杀生 (killing), 不与取 (stealing), 邪婬 (sexual misconduct) - 口业 (speech deeds, four): 妄言 (false speech), 两舌 (divisive speech), 麤言 (harsh speech), 绮语 (frivolous talk) - 意业 (mind deeds, at least two clearly named): 贪伺 (greedy watching / covetous scheming), 嫉恚 (jealous wrath)
The two listings overlap heavily but not perfectly. The *Ekottara* list includes the mental poisons of 贪他, 起恚, and 邪见 as numbered items, treating them as parallel to the physical and verbal actions. The *Madhyama* organizes everything by the three doors of action and includes 麤言 (harsh speech) as a distinct category — the quality of voice that wounds, that "辞气粗犷、恶声逆耳，使人苦恼而不得定" (rough in tone, harsh in sound, jarring to the ear, causing others distress and unable to bring them to stillness). Both lists are valid; they are slicing the same territory from different angles, the way you can sort books by color or by topic.
The *Madhyama* also gives classical definitions worth preserving verbatim because they are precise in ways our everyday language is not:
- 杀生 is defined as having *intention to harm* (有伤害欲) and *lack of compassion toward beings even down to insects* (对众生乃至昆虫不慈). It is not measured by outcome; it is measured by the heart behind the act. - 不与取 is defined as *being attached to another's property and taking it with the intention of theft* (贪著他人财物以偷盗之意取走). Possession of something another's hands gave you is not the same as taking it without their hands giving it. - 邪婬 is defined by the social situation of the person: those under the protection of parents or family, those already belonging to another, those whose violation carries legal penalty or fear (受父母、亲属守护，或已属他人、涉及惩罚恐怖). This is a relational and consent-based definition, not a list of forbidden acts in the abstract. - 妄言 has a famous four-form structure: 不见言见 (saying "I saw" when you did not), 见言不见 (saying "I didn't see" when you did), 不知言知 (saying "I know" when you don't), 知言不知 (saying "I don't know" when you do). And the motivation can be *for yourself* (为己), *for another* (为他), or *for money/property* (为财物). The same lie is still a lie whether you tell it for any of these reasons. - 绮语 is defined by five disqualifying marks: it is 非时 (untimely), 不真实 (untrue), 无义 (pointless), 非法 (against the Dharma), and 不能导向止息 (does not lead toward cessation). Any one of these is enough. Speech that scores well on all five is the opposite; speech that fails even one is 绮语. - 贪伺 is defined as *seeing another's wealth or means of livelihood and scheming to obtain it* (见他人财物及生活资具而伺求欲得). - 嫉恚 is defined as *harboring jealous hatred and arising with intentions to kill, bind, seize, or banish* (怀憎嫉而起杀、缚、收、逐擯等损害之念).
These definitions are doing real philosophical work. They are not vague feelings; they are structural features that can be observed.
The Roots and the Aggravating Factor
The texts also point to a root condition that makes unwholesome conduct more likely: 懈惰 (kausīdya — laziness, sloth). The *Ekottara* puts it as a fixed pairing: *"懈惰种不善行，于事有损"* — laziness plants unwholesome conduct and damages whatever it touches. The implication is that the *opposite* of laziness — 精进 (vīrya — diligence, energy) — is what plants wholesome conduct and benefits whatever it touches. So unwholesome conduct is not just a list of bad things; it is a list of bad things whose probability rises sharply when the mind is unwilling to look clearly, unwilling to apply itself, and willing to let things slide. This is important for the contemporary reading below.
The Continuity Into Future Lives
The *Madhyama* extends the picture across lives: *"前世行不善行，因此生不善漏令至后世"* — unwholesome conduct performed in a previous life produces what it calls 不善漏 (akuśala-āsrava — unwholesome outflows / corrupting tendencies) that continue into the next. This is not a threat; it is a description of momentum. Once a pattern of unwholesome action has been established in the mindstream, it continues to "leak" — to pull toward repetition — even into a new life. The practice the text offers against this is simply 身、口、意护 — guarding the three doors of action.
A Case That Seems To Bend The Rule
The *Ekottara* records the case of 鸯掘魔 (Aṅgulimāla), a serial killer who, in this very life, murdered a great number of beings and yet encountered the Buddha and attained the path. A contemporary reader immediately asks: does this case disprove the rule that unwholesome conduct ripens in suffering? The text's own way of handling this is important: it presents the case but does not retract the principle. The implication is that ripening is far more complex than a one-to-one ledger; it depends on conditions, on the depth of remorse and transformation, on the karmic field in which an act occurs. But none of this lets the rest of us off the hook. Aṅgulimāla is one case, and the text uses him as a question, not as a loophole.
Walking Through This in Daily Life
This is where the classical definitions either earn their keep or stay museum pieces. Let me walk through them one by one with contemporary scenes, and flag the places where contemporary people most often misidentify what the category actually covers.
杀生 (killing / taking life). Most contemporary people, on hearing this, immediately picture either homicide or vegetarianism. Both are partial. The classical definition is *intention to harm + lack of compassion toward beings even down to insects*. So this category includes: the bully at work who systematically dismantles a colleague's confidence over months (the being in question is still alive, but the intention and lack of compassion are identical in structure); the comment thread pile-on that drives someone offline; the clinician who speaks callously about a patient; the manager who fires someone in a way designed to humiliate rather than to communicate. On the other end, the category also includes a vegetarian who, when a mosquito lands on them, swats it with rage and a curse — that is also 杀生 with the marks the *Madhyama* names. The category does not care about your politics about meat. It cares about the quality of heart in the act.
不与取 (taking what is not given). Office supplies are obvious. Less obvious: taking credit for a colleague's work; using company time for your side business; lifting someone's design, code, or words and presenting them as your own; "borrowing" a friend's idea and pretending you came up with it; cutting in line; sneaking into a movie theater without paying; downloading pirated content; in relationships, taking emotional labor without reciprocal care. The classical definition is precise: being *attached to another's property* and *taking it with theft-intent*. That last word is key. If you genuinely don't know something isn't yours, it isn't 不与取. If you know perfectly well and take it anyway, it is.
邪婬 (sexual misconduct). The classical definition is relational and consent-based, not act-based. It flags violations involving those under protection of family, those already belonging to another, and situations involving fear of punishment. Contemporary translation: affairs with people who are married or otherwise partnered; relationships with subordinates where power dynamics cloud consent; sexual contact with those who are protected by family obligations and would be harmed by exposure; abuse of any kind. What it is *not*: a list of forbidden positions, or a blanket ban on sexuality. The category is about who is being harmed and how, not about what the body is doing in private.
妄言 (false speech). The four-form structure (saying you saw what you didn't, saying you didn't see what you did, saying you know what you don't, saying you don't know what you do) covers more than outright lies. It covers *the selective presentation of fact*. "I was working from home all day" (true, but you were also on Twitter for six hours); "We're like a family here" (false, in a way the boss means sincerely and the employees know is false); "We had no choice" (false, there were always other choices); the resume that inflates. And the motivations the *Madhyama* lists — for yourself, for another, or for money — are worth sitting with. The white lie told "to protect someone" is still 妄言 under the classical definition. That doesn't mean you should never tell a white lie; it means you should know what you're doing when you do.
两舌 (divisive speech). The classical definition: *transmitting words to break apart those who are harmonious*. Contemporary: telling one friend what another said about them, in a way designed to wound; "I probably shouldn't tell you this, but…"; sharing a screenshot of someone's private message with a third party; "Just so you know, she said…" Most contemporary people, when they do this, would describe themselves as *sharing information* or *being a good friend by keeping someone informed*. The classical category names what is actually happening: you are parting people who were together.
麤言 (harsh speech). This one doesn't appear in the *Ekottara* ten, but the *Madhyama* includes it because the Buddha cared about the *quality* of speech, not just its content. The classical description — rough in tone, harsh in sound, jarring to the ear, causing distress, unable to bring to stillness — describes the road-rage honk, the email written in all caps, the sarcastic comment made in front of a group, the voice that drips contempt. *What you say* matters; *how it lands in the ear* matters separately. Plenty of true things said in harsh speech become unwholesome precisely because of the harshness.
绮语 (frivolous, empty talk). The five disqualifying marks are a beautiful diagnostic tool. Untimely — a phone call at 2 AM about nothing. Untrue — gossip passed off as fact. Pointless — most doomscrolling sessions, most meetings that could have been emails. Against the Dharma — speech that encourages harmful action in others. Not leading toward cessation — speech that pulls you (and whoever is listening) further into the tangle rather than loosening it. The category is not "no fun allowed"; it is "speech that, on inspection, fails one or more of these five tests." Most casual conversation passes the test. Much of what fills social media does not.
鬪乱彼此 (inciting discord between people). The political operative who frames every issue as a battle; the influencer who builds an audience by pitting subgroups against each other; the manager who plays favorites to keep teams in competition; the family member who stirs old feuds at every gathering; the algorithmically optimized content designed to make you hate the other side. The category is structurally distinct from 两舌: 两舌 is telling something to break people apart; 鬪乱 is actively engineering the fight. The first gossips; the second agitates.
贪他 (covetousness). Wanting what belongs to others. Not admiring a beautiful house from afar — that is just noticing. Wanting their house, their spouse, their title, their follower count, their ease, in a way that bites. The classical definition is about *sizing up* another's goods and *scheming* toward them. The contemporary version is the comparison-driven scroll: looking at someone else's life and feeling the slow burn of "why not me," and from there to schemes (small or large) to obtain what they have.
起恚 (ill-will, wrath). The category names the *arising* of wrath, the volitional stirring up of harmful intent. There is a difference between the flash of anger that passes through you when someone cuts you off in traffic — that is an emotion arising in the mind — and the deliberate cultivation of resentment, the grudge you nurse, the schadenfreude you savor. The classical category names the second: the willed, sustained, enjoyed hostility.
邪见 (wrong view). This is the deepest of the ten and the most slippery. Wrong view, in the Buddhist technical sense, includes denial of karma, denial of the efficacy of action, denial that there is anything beyond immediate material consequences, the view that nothing matters, the view that cruelty is fine because nothing has any continuity. Contemporary versions: the confident assertion that "we only live once, so anything goes"; the cynic's mantra that everyone is corrupt and effort is for suckers; the conspiracy theory that absolves you of responsibility because *they* are the ones doing evil. Wrong view is dangerous because it operates upstream of the other nine — once you believe action doesn't matter, the rest become easier to rationalize. The *Ekottara* notes that those who hold wrong view are reborn in 边地 (borderlands / peripheral regions), 聋盲瘖哑 (deaf, blind, mute), 不闻正法 (not hearing the true Dharma). That is, wrong view doesn't just affect what you do; it affects what you are capable of hearing.
Now, where do contemporary people most commonly *mis*map these categories?
1. Treating the list as a checklist of "big" external sins. Most of us do not murder, steal, or commit adultery in any obvious sense, and we conclude the categories don't apply to us. But six of the ten are *speech*, and three are *mind*. The bulk of the categories operate in the average day.
2. Mistaking intention for outcome. "I didn't kill anyone, so I'm fine on 杀生." But the classical definition is built on the intention, the lack of compassion. If the intention is there and the outcome didn't happen because circumstances intervened, the unwholesome quality of the act is not retroactively removed.
3. Confusing 瞋 (anger as feeling) with 起恚 (wrath as volition). Feeling angry is not the same as deliberately cultivating ill-will. The category is about the second.
4. Treating 邪见 as "having unusual ideas." It is not about being heterodox; it is specifically about views that deny the relationship between action and consequence, that deny the possibility of liberation, or that hold the self to be permanent and supreme.
5. Forgetting that 懈惰 (sloth) is upstream. Most unwholesome conduct does not arise from grand malice. It arises from the unwillingness to be bothered, the lazy drift into the easiest available action. The classical framing — sloth plants unwholesome conduct — names this exactly.
Why Contemporary People Need This
A contemporary reader might reasonably ask: in a world where we can lie at planetary scale, where speech is instantaneous and permanent, where the most private acts are photographed, where karma's delays make it invisible to the liar — why bother with this ancient list?
Several reasons, and they are not sentimental.
The delay is not the absence of the cause. Modern life has trained us to expect feedback loops of milliseconds. If a button doesn't respond in 200ms, we call it broken. Unwholesome action operates on longer timescales — sometimes this life, sometimes the next — and the contemporary mind is structurally bad at handling long-delayed consequences. Climate change is a contemporary parable here: the cause is real, the effect is delayed, and the difficulty of getting people to act is precisely the difficulty of accepting delayed feedback. The Buddhist analysis of karma is, among other things, a long-term feedback-loop training. It says: just because the dashboard is silent does not mean the engine is not heating up.
Anonymity and scale have made the categories more relevant, not less. Killing in warfare is industrial now. Stealing happens by algorithm (insider trading, wage theft, systemic extraction). False speech is amplified by platforms. Divisive speech is a business model. The classical categories do not become obsolete in the face of these scales; they become *more* necessary, because the harms are now larger and the actors more diffuse. The texts cannot name drones or algorithmic radicalization, but the categories they describe — intention to harm, attachment to another's property, the four forms of false speech, the scheming to obtain, the engineering of discord — all apply, because the underlying mechanisms are the same.
The mental categories are more important than ever. The *Madhyama* classifies 意业 (mind deeds) as part of the structure, not as a footnote. In a contemporary environment that constantly rewards outrage, comparison, scheming, and contempt, the three mental categories — greed (贪伺), wrath (嫉恚), and wrong view (邪见) — are doing most of the heavy lifting of most people's suffering. They are also the categories least likely to be caught by any external accountability mechanism, because they leave no visible trace. The texts know this; they put them on the list for exactly this reason.
It is a tool for self-observation, not for judging others. This is the move the contemporary reader most needs to make. The list is not a lens for scanning other people. It is a mirror. When you sit with the ten categories at the end of a day, the question is not "who around me did what" but "which of these did *I* do, in which form, with what intention." The categories were given to trainees, not to judges.
It is a foundation for the rest of the path. Without a workable understanding of unwholesome conduct, meditation becomes an escape hatch and study becomes entertainment. With it, both become capable of doing what they are designed to do — loosen the hold of the patterns that bind the mind. Śīla (ethical training), samādhi (concentration), and prajñā (wisdom) are not three separate subjects; they are three aspects of one unfolding. Unwholesome conduct is what happens when the first aspect is neglected.
Common Misreadings and Clarifications
"This is just ancient morality, culturally specific." The categories are ancient; the mechanism they describe is not. Intention causing harm and harm ripeni ng in suffering is a description of how volitional action works, not a claim about which society one lives in. The specifics (which relationships count as protected, which speech is timely) are culturally inflected; the structural features (intention, harm, ripening) are not. A useful test: try describing a contemporary harm to someone and see whether you can do it without using any of the ten categories or their near-translations. You usually cannot.
"It means feeling guilty about everything." It does not. Guilt is often a way of avoiding the actual work of seeing and changing. The categories describe *deliberate, harmful action*. Slipping on ice and bumping into someone is not 杀生. Saying something clumsy that hurts a friend is not necessarily 麤言, especially if you correct it. The categories name *sustained, intended* patterns, not every passing ripple in the mind.
"Only really bad actions count." The *Madhyama* definition of 绮语 — speech that is untimely, untrue, pointless, against Dharma, or not leading to cessation — applies to a great deal of ordinary speech. The point of the list is not to identify the worst acts and ignore the rest; it is to develop a finer and finer discrimination, so that you can see and adjust at smaller and smaller scales. Practice over time increases sensitivity, not decreases it.
"The Aṅgulimāla case proves the rule doesn't work." It does not. The case is presented in the *Ekottara* precisely as a question — a tension to hold — not as a refutation. The text does not retract the principle, and it does not offer a clean theoretical resolution. What it offers is the reminder that ripening depends on conditions, including the most powerful condition of all: meeting a Buddha and undertaking the path. The case does not give the rest of us a loophole; it offers a complication that the wise student holds lightly.
"Following this list makes you passive." The opposite is the case. The texts pair 懈惰 (sloth) with the planting of unwholesome conduct, and 精进 (diligence) with the planting of wholesome conduct. The ethical training is not about withdrawal; it is about applying yourself more carefully to what you actually do. The list is a training program, not a resignation.
"It requires you to be perfectly pure before anything else matters." It does not. The texts say unwholesome conduct must be *故作* (deliberately done) to bear the full ripening. They also repeatedly offer practice in the very same breath as they identify the harm — *身、口、意护* (guarding the three doors). The training is the practice, and the practice begins wherever you are, not after you have become someone else.
A final note, because it is worth saying. This list has been around for a very long time, and the reason it has lasted is not that it is comfortable. It is that when a person actually sits with it — not to judge others, not to perform purity, but to see what they themselves are doing with their body, their mouth, and their mind — something shifts. Not all at once. Not dramatically. But the categories begin to function as a kind of soft instrumentation on the inner life, making visible what was previously invisible. And once you can see it, you can begin — with patience, and without despair — to do something other than keep doing it.
That, more than any vivid image of red-hot iron, is what the texts are pointing toward.