Madhyamaka and Yogācāra: Different Ways of Examining Experience

Epigraph

“…there is nothing that is not empty.”

Nāgārjuna, Fundamental Verses on the Middle Way, 24.19, concluding phrase. Editorial English translation of our Russian rendering of the English excerpt cited in Dan Arnold’s academic discussion, §2b; this wording is not presented as a verified published English translation.

About this article

You read a short message: “We need to talk.” Just four words, yet much has already happened inside you. Yesterday’s argument comes back, tension appears, and you think the relationship is ending. In this hypothetical example, the message really was received. But did it arrive with evidence of catastrophe? Where does the event end and the world that perception has already constructed begin?

Madhyamaka and Yogācāra are two influential currents of Mahāyāna Buddhist philosophy. Both investigate experience for the sake of liberation from ignorance and suffering, but offer different ways to test what seems obvious. Madhyamaka asks whether a phenomenon has an independent intrinsic nature. Yogācāra examines consciousness and how distinctions between perceiver and perceived arise. Their disagreement goes deeper than the familiar opposition between “the world exists” and “it is all in your head.” We will examine the necessary minimum of doctrine, then separately attempt a contemporary reading: noticing premature conclusions, taking actual circumstances seriously, and remaining responsible for action. This reading neither replaces the Buddhist path nor turns ancient philosophy into a scientifically validated psychological method.

I. Two investigations within a shared path

It helps to change the question at the outset. These schools did not develop as competing ways to become a more efficient employee or worry less about messages. Their task belongs to the Buddhist horizon of liberation: transforming our relation to experience so that the grasping that sustains suffering ceases.

Both currents developed in Indian Mahāyāna and subsequently influenced other Buddhist cultures. Nāgārjuna is especially associated with Madhyamaka; Asaṅga and Vasubandhu with Yogācāra. The names designate historically changing traditions of argument rather than two unchanging sets of answers. The academic overview of Madhyamaka introduces their relationship in this way.

The word “experience” itself needs care. To a contemporary reader it may mean only a private feeling: what I felt this morning. Buddhist philosophy considers much more—perception, action, causation, the habit of appropriating experience, and the possibility of liberation. Discussion of consciousness therefore cannot be reduced in advance to therapy, or discussion of emptiness to an exercise in positive thinking.

Consider another hypothetical difficulty. Someone is convinced: “I always fail at important conversations.” One investigation asks whether this unchanging failure exists independently of particular episodes, skills, interlocutors, and circumstances. Another examines how memory selects episodes, expectation directs attention, and attention makes the next mistake especially conspicuous. These are different questions. We can compare them without declaring them identical.

An important distinction already emerges: philosophy examines the grounds of habitual certainty, while our everyday example helps us feel the question. The example does not prove the doctrine. Even if revising a self-description brings relief, it does not follow that the person has established the truth of Buddhist metaphysics.

II. Madhyamaka: what absence of intrinsic nature means

Madhyamaka’s central term is śūnyatā, emptiness. It refers to the absence of svabhāva: an intrinsic nature through which something would exist by itself. Nāgārjuna investigates the dependence of phenomena and concepts; emptiness is not proposed as a special hidden substance. The discussion of Nāgārjuna connects this question with the way we fix the categories of experience.

We must not jump from “does not exist independently” to “does not exist at all.” Take an ordinary table. It depends on materials and manufacture, has parts, and serves a function in a particular situation. Remove it and there is nowhere to put the cup. Its dependence does not make the cup immune to falling. Yet no single leg is the whole table, and the word “table” does not create an object from nothing.

This is our explanatory example, not an account of a particular chapter of Nāgārjuna. It shows the difference between a functioning thing and a thing to which complete independence has been attributed. Sometimes that added independence creates the confusion: an object, person, or relationship supposedly must retain one essence under every condition.

This move is especially visible in relationships. “He is inattentive” may be a useful shorthand for repeated behavior. But the description can become a verdict: inattentiveness is declared the person’s ultimate nature, and new circumstances cease to count. Revising the verdict does not oblige us to tolerate unwanted behavior. It lets us say more precisely what happened, what recurs, and on what basis we decide.

Dependence can easily be turned into an endless excuse. Upbringing, fatigue, and environment influenced an action—therefore nobody is responsible for anything. That conclusion does not follow from dependence itself. If conditions matter, our actions matter as part of those conditions. The possibility of changing them becomes more intelligible, not less necessary.

III. Why emptiness does not abolish suffering or truth

The strongest test of a doctrine is an objection, not admiration for its paradox. If everything is empty, why can we speak of suffering, a path, and liberation at all? In chapter twenty-four of the Fundamental Verses on the Middle Way, Nāgārjuna answers a similar objection. In the academic explanation of this argument, dependence is a condition of change: something completely unchanging by nature could not be brought to an end.

A contemporary reader can explore the point through an example of our own. A hypothetical person says: “Anxiety is simply who I am.” If this were literally an unchanging essence, differences between morning and evening, rest and overload, solitude and support would have to be irrelevant. Yet the person notices differences. That already makes it more useful to investigate what is happening than to declare a state a complete definition of oneself.

We cannot promise, however, that this investigation will remove a severe condition. A philosophical argument and help for a particular person are different things. Circumstances may require professional support, a changed environment, or prolonged work. Asking about the absence of an independent essence establishes neither a cause nor the appropriate form of help.

The distinction between conventional and ultimate truth does not authorize lying either. In everyday life there is a difference between the right address and a wrong one, a kept promise and a broken one. Investigating the ultimate grounds of concepts does not erase that difference. Otherwise philosophical language would merely conceal a refusal to check facts.

We can express the distinction through two disagreements that appear to concern the same object. In our hypothetical example, someone disputes a debt because it has already been paid. Resolving this requires information about the payment and agreement. Another question concerns whether the debt exists independently of participants, rules, and previous actions. Pointing to that dependence does not tell us whether payment occurred. These are different ways of changing the original assertion: correcting a fact or reconsidering an assumed mode of existence. Confuse them, and “empty” starts doing another question’s work, as though it could settle a dispute about debt by itself. Distinguish them, and examining foundations does not obstruct establishing facts. Indeed, it becomes clearer which objection is relevant: evidence of a money transfer may change the conclusion about performance of an obligation, while a philosophical argument about dependence cannot replace that evidence. The analogy clarifies the distinction between questions; it is not an independent proof of Madhyamaka.

We might put it this way: a short description is necessary for action, but need not become the last truth about the world. We may use “house,” “agreement,” and “I” without asserting that every designation discovers a separate eternal essence. This discipline of language is especially valuable where one successful label begins to replace investigation.

IV. Yogācāra: experience arrives already organized

Yogācāra considers how consciousness participates in the appearance of the experienced world. Cittamātra and vijñaptimātra are often rendered “mind-only” and “representation-only.” Their philosophical meaning is disputed: some scholars see strong idealism, while others emphasize analysis of access to experience. The overview of Yogācāra shows the internal diversity of texts and interpretations.

We need not immediately settle the existence of an external world to enter the problem. It is enough to notice that perception does not resemble the delivery of fully finished meaning. There are two phrases on a screen, yet a person encounters a threat, request, command, or invitation. A transition occurs from signs to a situation. It may be reasonable, mistaken, partly checked, or still unknown.

Take a hypothetical workplace conversation. A manager says: “We will return to this tomorrow.” One employee hears rejection. Another hears acknowledgment of the task’s importance. A third notices only a calendar deadline. Different readings do not mean they are all true, or destroy the possibility of clarifying the manager’s intention. They show that personal certainty already contains the work of perception.

“Consciousness participates” is nevertheless too weak a formula to exhaust historical Yogācāra. It works as a contemporary entry point, but some texts defend a much more radical position. Reducing the school to the banal observation that people sometimes make mistakes makes it convenient at the cost of its substance.

Both sides must be retained. We should neither advertise the doctrine as a proven fact of neuroscience nor neutralize it until only the familiar advice to check guesses remains. The philosophical difficulty lies between these simplifications: how do we explain the structure of experience, and what is the object given to consciousness? Even a reader who rejects the proposed answer can recognize the seriousness of the question.

The teaching of the three natures makes the analysis more precise. In Vasubandhu’s account, the imagined nature concerns how experience is taken to be the existence of an independent object and a perceiving self; the dependent nature concerns the conditioned process in which that experience arises; the perfected nature concerns the absence of the attributed existence and duality. The SEP discussion of the three natures shows that these are connected aspects, not three floors of the world with another substance concealed on the top floor.

Our comparison with Madhyamaka now yields a more definite question. Showing that attributed independence does not withstand analysis is one task. Explaining how the attribution arises, what conditions sustain it, and what changes when the error ceases is another. These tasks may intersect, but their formulation and philosophical vocabulary differ.

The perfected nature therefore must not become an attractive name for my “true self.” That conclusion reinstates an independent essence precisely where its absence is being examined. Nor should the dependent nature be understood merely as acknowledgment that thoughts affect mood. Yogācāra’s soteriological question concerns the deep organization of experience and liberation, not only a convenient change of everyday explanation.

When reading, we need to keep both sides together. Arising experience need not exist in the way we take it to exist; the absence of attributed existence does not mean the absence of a conditioned process. Lose the first side, and criticism becomes a new metaphysical foundation. Lose the second, and it becomes an empty denial of everything happening. This distinction helps explain why the schools’ disagreement requires more than a choice between the words “world” and “consciousness.”

V. Habits of perception without an eternal soul beneath them

Why does yesterday’s habit influence today’s experience if everything changes? Yogācāra offers the language of “seeds” and ālaya-vijñāna, often translated as storehouse consciousness. In the overview of Vasubandhu, this model explains continuity and delayed consequences of actions without introducing a permanent ātman.

The concept cannot simply be equated with the unconscious in contemporary psychology or called an ancient description of brain memory. The model belongs to its religious and philosophical context, including karma and the path of liberation. Similar words—“trace,” “habit,” “memory”—do not establish identical theories.

Our contemporary observation is much more modest. If someone has repeatedly expected humiliation in arguments, a new argument may quickly assume a familiar pattern. They do not necessarily choose this reading intentionally. But we cannot turn the observation into a final explanation of their biography either: perhaps the current interlocutor really is humiliating them. A history of perception and a present situation may operate simultaneously.

This matters fundamentally for responsibility. A person may participate in constructing meaning without being to blame for every harm inflicted on them. Conversely, a habitual reaction does not mean that every cruel response of one’s own can be called automatic and therefore inconsequential. There is room for precise examination between blaming oneself for everything and denying any participation.

The practical question here differs from “What is my hidden essence?” It asks: what do I notice first, what do I miss, what explanation arises before checking, and what action usually follows? The answer remains revisable. A single episode does not prove a universal pattern, and a discovered recurrence does not become destiny.

VI. A disagreement that an attractive synthesis cannot erase

Can Madhyamaka and Yogācāra be considered two halves of one doctrine? Comparing them is useful; declaring them reconciled in advance would be inaccurate. Their dispute concerns, among other things, the role consciousness may have in explaining reality, and whether an explanatory foundation becomes a new independent essence. Their history includes both criticism and attempts to combine approaches.

To feel the problem, let us use an analogy of our own. One investigator notices that a map depends on purposes, scale, and rules of representation. Another studies the mapmaking process in detail. The second may answer questions the first left open. But the first can then ask: have you declared the mapmaking process itself completely independent and an ultimate foundation?

This analogy says nothing about how a particular Buddhist author would answer every question. It helps distinguish two intellectual tasks. Describing a mechanism does not always settle the question of its status. Criticizing an independent foundation does not always explain how a particular experience arises.

A popular synthesis often springs from the desire to obtain a worldview immediately, without inconvenient disagreements. “They all said the same thing” sounds peaceful. Yet respect for a tradition sometimes requires recognizing that its authors regarded a disagreement as substantial. Otherwise we retain their names while replacing their questions with our answers.

A contemporary reader need not choose a winner. We can investigate what each approach makes us notice, where disagreement arises, and which assumptions seem persuasive. This openness differs from arbitrary mixing. In the first case, boundaries are named; in the second, their absence is presented as higher wisdom.

Disagreement has another virtue: it separates a question from the need to secure agreement immediately. Contemporary readers are accustomed to looking for a position without contradictions, then using it as a universal key. Human situations, however, often need several descriptions. A conversation can be examined as an exchange of words, a history of expectations, a relation of power, and an accepted obligation. No description automatically has to eliminate the others.

Suppose that in a hypothetical family one person repeatedly promises help and fails to provide it. Another notices a painful reaction to uncertainty in themselves. These facts are compatible. It would be wrong to say that because the reaction has a history, the broken promise has become an illusion. It would also be wrong to say that because the promise was broken, there is nothing left to investigate in one’s perception. Different questions preserve the complexity needed for a decision.

Several descriptions do not mean endless reflection is required. Sometimes enough is already known to act. We know that an agreement was broken, that repetition causes specific harm, and that the previous way of discussing it did not help. Further investigation may refine action, but need not postpone it until complete philosophical clarity. Responsibility exists amid unfinished explanations.

Examining certainty should also distinguish doubt from helplessness. “My explanation may be incomplete” leaves room to ask and correct. “Nothing can be known” closes that room. The first assertion has a practical continuation: what information would change the conclusion? The second has none, which is why it sometimes offers convenient protection from an unwelcome fact.

This distinction helps us read difficult traditions without consumer haste. We may fail to understand an argument on the first reading, refrain from choosing a school, and decline to turn every idea into an immediate tool. Slow reading can also be a useful action. It teaches us to sustain a question without replacing it with the answer most easily fitted into our familiar picture of the world.

VII. Three attractive mistakes

The first mistake is to turn emptiness into an emotional command: “Nothing exists, so stop worrying.” Philosophy then becomes a means of suppressing a person. But the feeling already exists as an event of experience. Investigating it does not begin by demanding its disappearance.

The second is to turn Yogācāra into a slogan claiming that an individual thought creates everything at will. It is easy to proceed from this to blaming someone who was harmed: supposedly they “created” the injustice themselves. Neither Buddhist analysis of consciousness nor philosophical idealism alone warrants that conclusion. We must separately explain why the event arose, who acted, what can be checked, and what measures are available.

The third is to find an eternal unified self beneath all differences and call it the Buddhist conclusion. That combination may belong to someone’s contemporary worldview, but cannot be attributed to these schools without explanation. Shared words about depth and unity do not erase the Buddhist question of not-self.

There is a subtler mistake too: deploying doubt only against unwelcome conclusions. When praise feels good, we take it literally; when criticism hurts, we declare everything conventional. The philosophical tool then serves the same old grasping. A substantive examination must allow revision of flattering descriptions as well as frightening ones.

For example, “I am completely indispensable” can keep someone overloaded more effectively than “I am not good enough.” Both assertions can conceal conditions and particular obligations beneath the enormous word “I.” We can check the division of work, agreements, actual consequences of refusal, and available help. Neither pride nor fear replaces this information.

VIII. A similar question about the self can lead to opposite foundations

Philosophies of experience should be compared through what their question does to the foundations of knowledge. Similar doubt does not guarantee a similar answer. This is especially clear when the perceiving self is examined: it may be sought as a reliable support, explained as a sequence, or approached by questioning the very need for an independent support.

For Descartes, doubt leads to certainty that the thinker exists in the act of thinking itself. The academic discussion of the cogito, §4.1, clarifies that this first step does not yet assume a complete metaphysics of a substantial self. Immediately opposing every sentence of his to Buddhist not-self would therefore be inaccurate. But the investigations have different directions. Descartes seeks a foundation able to withstand doubt; Madhyamaka questions the independent status of supposed foundations. What they share is seriousness of examination, not a discovered ultimate entity.

Hume offers a different comparison. In the discussion of his account of personal identity, a permanent self is not found as a separate impression; a bundle of perceptions is discussed, although Hume himself notes a difficulty with his explanation. This helps raise the question of continuity without a ready-made owner of experience. Similarity of formulation, however, does not turn Hume’s sequence of perceptions into ālaya-vijñāna. Karma, liberation, and Buddhist investigation of duality give concepts a different task. Agreement in a negative step does not yet provide a shared positive doctrine.

Finally, Spinoza shows the limit of the popular slogan of universal interdependence. His finite modes exist in one necessary substance, as explained in the SEP account of Ethics I. Madhyamaka’s emptiness must not be turned into such an ultimate substance. Both approaches constrain the picture of isolated finite existence, but answer differently what makes dependence possible.

This distinction matters when reading the original message. We can examine certainty, the history of perception, and the grounds of words without promising that every inquiry will lead to one worldview. Comparison broadens the question. It does not present a feeling of recognition as evidence of agreement or substitute universal psychology for the Buddhist path.

IX. Conclusion: clarity need not become the last word

Madhyamaka and Yogācāra are compelling not because they offer a quick way to conquer an unpleasant feeling. They question the ease with which experience appears already understood for good. A confident name for a thing, a habitual self-image, and another person’s instantly interpreted intention may contain more assumptions than first meet the eye.

Different questions retain different tasks. Testing an independent essence weakens the conversion of a temporary description into a final verdict. Investigating the construction of experience shows consciousness participating in what appears immediately given. Neither move removes the need to check circumstances or responsibility toward others.

A reader who does not accept Buddhist religious premises still has room for honest discussion. We can investigate an argument without claiming awakening. We can acknowledge the complexity of perception without denying our neighbors’ existence. We can attend more carefully to words without abandoning promises and boundaries.

Return to the few words on the screen. Perhaps the conversation will be difficult. Perhaps it will concern something entirely different. While the distinction between the message and a finished story remains visible, the person has a small space for checking and choosing. This is not yet liberation in the Buddhist sense. But it is already a more precise engagement with the experience actually occurring.

What this means in practice

What follows is a contemporary clarity exercise inspired by the distinction between questions, not a traditional Yogācāra or Madhyamaka practice. Choose a small episode in which a decision can be reconsidered without much risk. Do not begin with an event that currently feels too difficult to investigate.

First describe what happened in words another participant could check. “The reply came two days later” is more precise than “I was rejected again.” Then record your own interpretation and feeling. They matter, but belong to different layers of description.

Now apply the first question: on what conditions does your definition of the situation depend? If it is “the relationship is destroyed,” what events support it, what could change it, and what remains unknown? You need not obtain a comforting version. Sometimes checking really does reveal a serious breach of agreements.

Then apply the second question: how did this particular reading form? What did you notice first, what memory appeared, and what was added to the literal words? Do not call every reaction an error. If past experience helped you notice danger in time, that too needs to be retained.

In the hypothetical message example, a reasonable step might be clarifying the conversation’s subject. If there is already a clear recurring breach, the step might be stating a boundary. If information is missing, leaving the conclusion open is permissible. The exercise ends not in compulsory calm, but in a more differentiated picture: event, interpretation, unknowns, possible action.

After acting, return to the original description. Was the assumption confirmed? Did understanding change? What happened independently of expectations? This return check protects reflection from becoming an attractive story that never encounters circumstances.

Sometimes intention cannot be checked. It then helps to move from inaccessible knowledge to an available choice. You can name the deadline by which you need a reply, request a written agreement, or decline an obligation whose conditions remain unclear. This is neither victory over another person nor a demonstration of philosophical superiority. It is a way to act without presenting an assumption as an established fact.

Finally, the exercise should not be used to reform an interlocutor. “You are just projecting” usually checks nothing, but readily deprives someone of the right to describe their experience. We can propose clarifying an event together. We cannot appoint ourselves in advance as the one who sees reality, and another as someone made entirely of illusions. Even an excellent philosophical vocabulary confers no such authority.

Returning to the message now means checking one accessible question rather than constructing a final theory of the interlocutor. Here the two philosophical approaches retain their different tasks: examining a fixed foundation and examining how a reading emerges.

Key quotation

“What appears is dependent. How it appears is imagined. Because it depends on conditions. Because it is merely a fabrication.”

Vasubandhu, Exposition of the Three Natures, verse 2 in full; editorial English translation of our Russian rendering of the English version in SEP, §4.3, not a verified published English quotation. The verse connects the appearance of experience with dependence on conditions and distinguishes it from an attributed mode of existence. The question therefore ends neither in affirming an unchanging object nor in denying appearance itself.

Connections across traditions

These are our analytical comparisons. The table shows a shared question and the limits of similarity; agreement of formulations alone does not establish historical influence.

Positions	Similarity	Boundary
Madhyamaka ↔ Spinoza	The finite is not regarded as completely isolated.	Spinoza has one substance; emptiness does not introduce that foundation.
Yogācāra ↔ Epictetus, §5	Our reading of events needs examination.	Stoic judgment is not identical to the Buddhist model of an object’s appearance.
Ālaya-vijñāna ↔ Hume	Continuity without an unchanging owner becomes a question.	Karmic seeds are not Hume’s bundle of impressions.
Madhyamaka ↔ Descartes	What seems obvious undergoes philosophical examination.	Testing independent foundations and seeking the certainty of the cogito have different tasks.
Vasubandhu’s three natures ↔ Nāgārjuna	A phenomenon need not exist as we take it to exist.	Connected aspects of appearance do not replace the argument against intrinsic nature.

Sources

Vasubandhu. Stanford Encyclopedia of Philosophy. §4.3: the three natures and nonduality.
Richard P. Hayes. Madhyamaka. Stanford Encyclopedia of Philosophy. Introduction, §1, §§3–5: the school’s status, emptiness, and differences in methods of argument.
Douglas Berger. Nagarjuna. Internet Encyclopedia of Philosophy. Introduction and philosophical account: intrinsic nature, dependence, and Nāgārjuna’s place.
Dan Arnold. Madhyamaka Buddhist Philosophy. Internet Encyclopedia of Philosophy. §§1–3, especially §2c: the nihilism objection and the argument of MMK chapter 24; the primary text was read here through a limited academic discussion.
Szilvia Szanyi. Yogācāra. Stanford Encyclopedia of Philosophy. §§1–3: textual history, models of consciousness, the three natures, and contemporary disagreement about idealism.
Vasubandhu. Internet Encyclopedia of Philosophy. §6: storehouse consciousness, the Twenty Verses and Thirty Verses.
Descartes’ Epistemology. Stanford Encyclopedia of Philosophy. Academic discussion of doubt and the cogito..
David Hume. Stanford Encyclopedia of Philosophy. Academic discussion of personal identity and the difficulty in the Appendix to the Treatise..
Baruch Spinoza. Stanford Encyclopedia of Philosophy. Academic account of one substance and finite modes..
Epictetus. Enchiridion, §§1, 5, 51, translated by Elizabeth Carter. Primary text: impressions, judgment, and putting philosophy into practice..

About the authors

Andrey Zarinsh is the author and founder of United Wisdom. He explores inner transformation, the unity of traditions, intercultural dialogue, and contemporary spiritual discipline. Instagram: @andrey.zarinsh.

Maria Evilkut is a coauthor of United Wisdom. Alongside her research, she seeks to test these ideas in practice and observe how principles of unity and mutual understanding work in real life between people and cultures. Instagram: @ilkutki.