Who is the Doer? — The Agent Criminal Law Cannot Find

Law’s Invisible Assumption

#

Before any court can punish, it has to do something so basic it is rarely even named: it has to locate a doer. A specific person did a specific thing with a specific mental state. The whole machine of criminal justice — actus reus, mens rea, desert, punishment — is built on that single attribution. Law calls the question solved. Philosophy cannot.

But law never asks the prior question: what is a doer? What makes this person — rather than their conditioning, their neurology, their upbringing, their karmic inheritance, their social formation — the genuine author of the act? Philosophers call this the problem of agency. Law behaves as if it is settled. It is not.

Post 1 established that the Self of Advaita — pure witness-consciousness, the sakshi — is not the “person” law has in mind. It is unacting, unattached, and was never the originator of anything. Post 2 established that the freedom law presupposes at the moment of action is already inside a condition every Advaita tradition calls bondage — it is conditioned, beginningless, and never what retributive punishment requires. This post reaches the sharpest question the series has posed:

If the Self does not act, and bondage conditions every choice, and the ahamkāra is a construction — who, then, is the actual doer of any act? And what is criminal law punishing when it convicts someone?

Advaita offers three distinct and incompatible answers:

• Ashtavakra: No one has ever done anything. The Self is akartā — the non-doer — by nature. The ahamkāra appears to act but has no ultimate standing. The entire structure of doer, deed, and done-to is a superimposition in awareness.
• Shankara (Vivekachudamani): The jīva acts — but not as a free, self-originating first cause. It acts as a conditioned node through which the three guṇas of prakṛti operate. Responsibility is real within vyavahāra (the transactional order) but it is conditioned responsibility, not libertarian freedom.
• Gaudapada (Mandukya Kārikā 3): “Who is the doer?” is a question formed inside the dream. It has intra-dream answers. From paramārtha — the highest standpoint — neither doer, deed, nor victim ever truly arises. The question is not difficult. It is ill-formed.

This post does not close that collision. It holds it open — and shows exactly what it costs jurisprudence to continue looking away.

The Sanskrit of Doing — Why Precision Matters Here

#

The three texts work with a small set of technical terms that map almost perfectly onto what criminal law needs from an accused person — but they use those terms to say something law has never considered. Understanding this vocabulary is not preliminary. It is the argument.

Kartā — the doer, the agent, from the root kṛ (to do, to make). In Sanskrit grammar, the subject of a transitive verb. In metaphysics, the supposed originator of action. In court: the accused is treated as a kartā. Law cannot convict without one.

Bhoktā — the experiencer, the enjoyer, the sufferer. The one who undergoes the fruits (phala) of action. The convicted person serving a sentence is a bhoktā. Law’s punishment is addressed to the bhoktā.

Kriyā — action as pure event. The arm moves, the words are spoken, the blow is struck. Distinct from karma, which is action plus its moral and karmic loading, action as something that generates residue and consequence across time.

Akartā — the non-doer. This is the crucial term. Not “someone who refuses to act” or “someone who has achieved detachment.” But that which was never constituted as an agent — that which, by its nature, does not initiate, does not own, does not suffer consequences. This is what Advaita says the real Self is.

Ahaṃkāra — the I-maker. The mental mechanism that says “I did this” and “this happened to me.” It claims authorship retroactively. It is, functionally, the legal person — the entity courts address when they read charges, deliver verdicts, and impose sentences. And it is a construction. It arose from prior conditions. It has no independent existence beneath those conditions.

Guṇas — the three qualities of prakṛti: sattva (clarity, illumination, harmony), rajas (activity, passion, agitation), tamas (inertia, dullness, resistance). For Shankara and the Bhagavad Gītā, it is the guṇas that do the acting, not the Self. The Self is witness. The guṇas are the engine.

The mapping to law is precise and devastating:

Advaita’s claim is not that nothing happens. It is that the real Self is akartā; the apparent kartā is ahaṃkāra; and the actual causal engine behind every act is the guṇas — not a metaphysically free inner will standing outside the causal order and initiating from sovereignty.

Ashtavakra — No One Has Ever Done Anything

#

Ashtavakra’s position on doership is absolute and it does not qualify itself. The Ashtavakra Gita speaks at the level of someone already capable of letting the whole idea of “I act” fall away — not as a discipline, not as a gradual achievement, but as an immediate recognition of what was always already true.

Verse 1.6 — The Mind Is Not Mine

#

nāhaṃ deho na me deho bodho’hamiti niścayam na me rāgādayo doṣāḥ kriyākarma na me manaḥ

Transliteration: nāham deho na me deho bodho’ham iti niścayam / na me rāgādayo doṣāḥ kriyā-karma na me manaḥ //

Translation: “I am not the body, nor is the body mine. I am awareness — this is my certainty. I have no attachment or other defects. I have no action, no karma, and the mind is not mine.”

The jurisprudential force is in two phrases:

kriyā-karma na me — “action and karma are not mine.” Not “I am working toward non-attachment from my karma.” Not “I am purifying my actions.” Simply: they were never mine. The act did not originate here.

manaḥ na me — “the mind is not mine.” This is the direct assault on mens rea. Criminal law requires a mind that belongs to the accused — a mind that formed an intention, that knew what it was doing, that chose. Ashtavakra says: at the level of the real Self, even mind is not “mine.” The guilty mind belongs to the ahaṃkāra, not to the sakshi. Law’s entire subjective element of liability attaches to a case of mistaken identity.

This is not a plea of irresponsibility. It is a claim about the structure of reality. Bodies move. Minds form intentions. Stories of “I did this” arise. But the real subject — awareness itself, the witness — never becomes an agent. It watches the whole drama without becoming any part of it.

Verses 1.12–1.15 — The Already-Free Self

#

Ashtavakra moves from the negation directly to the portrait of what the Self actually is:

1.14: niḥsaṅgo niṣkriyo’si tvaṃ svaprakāśo nirañjanaḥ

“You are unattached, actionless, self-luminous, pure.”

The term asaṅga (unattached) is doing precise work. Not “detached after effort.” Not “becoming non-attached.” Already, by nature, unattached. Attachment belongs to the ahaṃkāra. The witness was never attached to begin with.

1.15: ayam eva hi te bandhaḥ samādhim anutiṣṭhasi

“Your bondage is only this: that you practise concentration.”

This is where Ashtavakra becomes most subversive. If you are already free and already akartā, then any practice that presupposes you are not yet free reaffirms the premise “I am an incomplete being who must become something.” That premise is the bondage. The practice entrenches exactly what it claims to dissolve. The one who thinks they are a doer who must undo doership has doubled down on the original error.

For jurisprudence: every rehabilitation program that tells a person “you are broken and must be fixed” may be doing precisely what Ashtavakra warns against — reinforcing identification with the role of “criminal self” that it claims to correct.

Verse 15.4 — The Most Legally Explosive Statement in the Tradition

#

na kartā na bhoktā’si na jīvāvena mucyase prabuddho’si sadā śuddho niḥsaṅgo dvaitavarjitaḥ

Translation: “You are not the doer, not the experiencer. You are not a jīva being liberated. You are ever-awakened, ever-pure, unattached, free from duality.”

Two words carry the weight of an entire jurisprudential crisis:

Na kartā — not the doer. Unqualified. Not “not ultimately the doer.” Not “not the doer in the highest sense.” Not the doer.

Na bhoktā — not the experiencer, not the one who undergoes consequences. The one who serves the sentence is not the real Self. The one who is punished is not ultimately there in the way punishment requires.

The implication is precise: the real “you” has never committed an offence and has never suffered punishment. The entire legal drama — offence, investigation, trial, verdict, sentence, prison — unfolds between an apparent doer (ahaṃkāra claiming “I did this”) and an apparent sufferer (ahaṃkāra experiencing the consequences) in the field of appearance. The witness simply sees. It was never implicated. It cannot be reached by any sentence a court can impose.

Chapter 18 — The Jīvanmukta: Acting Without Being a Doer

#

Ashtavakra’s liberated sage — the jīvanmukta — still appears to act. Janaka walks, speaks, governs, eats, responds. Chapter 18 describes this precisely: the liberated one “does whatever gets done,” whose “life proceeds without him as the author.”

The paradox is deliberate and important:

• From outside, the jīvanmukta is a kartā. They act. Law would address them as an actor.
• From inside, there is no sense of “I do this.” There is only seeing actions happen. The witness watches the body and mind move without claiming it.

This sets a hard epistemological limit for law: law can only ever see the outside. It can observe behaviour. It cannot perceive akartā. All legal attributions of doership are necessarily operating at the level of appearance — the level Ashtavakra calls the transactional, the level of māyā. Law is not wrong to operate there. But it must know that is where it operates.

What Ashtavakra Means by Non-Doership — Three Distinct Arguments

#

These are not the same argument repeated. They are three independent lines of inquiry that converge on the same conclusion.

A. The Epistemic Argument

Whatever you observe — your body acting, your mind forming an intention, even the thought “I did this” — is an object in awareness. An object appearing to the subject. The subject is that which observes all of these. Now: the thought “I am the doer” is itself observable. It is an object. Therefore the “I” that is claiming doership is not the subject — it is itself an object appearing in awareness. The real subject does not say “I did this.” It watches “I did this” arise and pass. The witness is never the witnessed.

B. The Phenomenological Argument

Pay close attention to the actual sequence of any act. Any act at all. The movement begins. Awareness notices the movement. The narrative “I decided to do that” appears — after the movement, not before it, not as its initiating cause. The doership claim is retrospective. It is a story the mind tells, not a report from inside the causal engine. This is what Libet’s experiments later demonstrated neurologically. Ashtavakra reached the same conclusion through meditative introspection fifteen centuries earlier.

C. The Ontological Argument

Even granting that ahaṃkāra does act and does form intentions — it is a conditioned construction in awareness. It arose from prior conditions: past saṃskāras, the accumulated weight of previous identifications, the guṇa-composition of this moment. It has no independent existence beneath those conditions. It is like a wave. You can trace the wave. You can observe its behaviour. But you cannot hold the wave metaphysically responsible for what the ocean is doing. The wave is the ocean in a particular configuration. Punishing the wave for being a wave is not punishment in any philosophically defensible sense.

The consequence for law: responsibility exists only in the transactional order. Deterrence, rehabilitation, incapacitation — these all operate on the ahaṃkāra and can be more or less effective at that level. But retribution — the claim that the real subject of experience deserves to suffer in proportion to what it did — projects onto the Self a moral weight the Self cannot carry, because the Self was never the doer.

The Paradox of the Teaching Itself

#

If no one is the doer, who is Ashtavakra instructing? If Janaka is not a doer, why does Ashtavakra offer him teaching? If the ego is a fiction, why does the fiction need to hear anything?

The text’s own structure is the answer: the teaching operates at the level of appearance, for the one still identified with appearance. It is medicine prescribed in a dream for a dream-illness. When the patient wakes, both the illness and the medicine belonged only to the dream. The teaching is real within the transactional level — it produces real recognition, real shift. From the paramārthika level, there was no ignorance to be removed and no one who removed it.

Law is in an exactly analogous position. It addresses people as if they are free agents — because within the transactional frame, that is the only way to coordinate behaviour, prevent harm, and maintain a workable social order. That “as if” is not dishonesty. It is pragmatic necessity. The teaching is necessary. Law is necessary.

But when jurisprudence starts treating its “as if” model as a statement of ultimate metaphysical truth about the self — when it says not just “we will hold you responsible” but “you are ultimately, metaphysically, fundamentally responsible, and therefore you deserve this suffering” — it has stepped decisively out of its lane. It has confused the transactional level with the paramārthika level. And that confusion is where retributive punishment finds its most serious philosophical problem.

Vivekachudamani — The Jīva as Conditioned Doer

#

Shankara’s Vivekachudamani accepts non-doership at the level of the Self with complete agreement. Where it disagrees with Ashtavakra — sharply and importantly — is on whether the jīva (the individual soul) is a real doer within vyavahāra, and on whether bondage is a simple cognitive error correctable by recognition or a structural, beginningless condition requiring methodical dismantling.

For Shankara: ahaṃkāra is a real enough doer in vyavahāra, and it operates under the influence of the guṇas and beginningless avidyā. The jīva is responsible — genuinely, not merely fictionally. But its responsibility is conditioned responsibility, not the libertarian free choice retributive punishment requires.

Verse 104 — Egoism as Doer

#

Around verse 104, Shankara identifies the culprit with precise Vedic anatomy:

“Know that it is Egoism which, identifying itself with the body, becomes the doer or enjoyer and, in conjunction with the guṇas such as sattva, assumes the three different states — waking, dream, and deep sleep.”

The doer is precisely located: in ahaṃkāra, not in the Self. But with a crucial addition: doership only arises “in conjunction with the guṇas.” The ahaṃkāra does not generate action independently. It acts through and as a particular configuration of the three guṇas. The guṇas are the engine. The ego is the one who then says “I drove.”

Law’s “person” is this entity: ego identified with body, acting under guṇa-conditions. Shankara’s subtlety is that he treats this as real enough for karma and dharma — real enough to accumulate consequences, real enough to bear responsibility within vyavahāra — while insisting that from the Self’s side it is pure superimposition.

Bhagavad Gītā 3.27 — The Decisive Verse (Cited by Shankara Throughout)

#

prakṛteḥ kriyamāṇāni guṇaiḥ karmāṇi sarvaśaḥ ahaṃkāra-vimūḍhātmā kartāham iti manyate

Translation: “All actions are performed, in every case, by the guṇas of prakṛti. The one whose self is deluded by egoism thinks ‘I am the doer.’”

This is the pivot on which Shankara’s entire theory of action turns, and it is almost a compressed version of the jurisprudential argument of this post. The real causal story of any act — any act, including any criminal act — is:

1. A particular guṇa-composition is present in the mind and body at this moment: a specific configuration of sattva, rajas, and tamas.
2. Under that configuration, certain impulses, fears, desires, and aversions arise.
3. Action occurs — the body moves, words are spoken, harm is done.
4. Ahaṃkāra, deluded by avidyā, retrospectively claims “I did that.”

Shankara does not deny responsibility. He locates it: responsibility attaches to the identification — to the ego’s claiming of the guṇas’ action as its own. That identification is real. It has real karmic consequences. But the underlying agent — the actual causal engine — is never the Self. It is Prakṛti operating through its three constituents.

Gītā 3.28 — Guṇas Acting Upon Guṇas

#

tattva-vit tu mahā-bāho guṇa-karma-vibhāgayoḥ guṇā guṇeṣu vartanta iti matvā na sajjate

Translation: “But the one who knows the truth about the division of the guṇas and actions understands that it is only guṇas acting upon guṇas — and is not attached.”

“Guṇas acting upon guṇas.” This is the most sophisticated description of action in any philosophical literature. It means: what we call an “act” is one configuration of prakṛti (the acting mind-body system) impacting another configuration of prakṛti (the world). The wise see this and withdraw the “I” from the picture. They act — fully, effectively — without the claim of ultimate authorship.

This is the model law does not use. Law sees guṇas-acting-on-guṇas (neural causation, social causation, psychological causation) and adds a heavy metaphysical “I” as the sovereign origin of all of it. The Gītā strips that “I” back out — without denying that, within the field, actions are real and consequences are real.

Verses 254–265 — The Jīvanmukta: Acting Without Accruing Karma

#

Shankara’s liberated one — the jīvanmukta — continues to act. They speak, walk, eat, teach. But they no longer identify as the doer. The actions arise through the momentum of prārabdha karma — the karma already set in motion that gave rise to this embodiment and which must exhaust itself. But no new karma is generated, because there is no fresh identification of “I” with the guṇa-driven act.

The profound legal implication: law has no access to this interior shift. From the outside, the jīvanmukta and the ordinary conditioned person look identical — both act, both speak, both engage with the world. From inside, one is kartā and one is akartā. Law, which can only ever see the outside, cannot distinguish them. All its attributions of doership are therefore necessarily at the level of appearance.

Shankara’s Full Theory of Action — Guṇas, Not Persons

#

To understand what Shankara’s account of doership actually demands of criminal law, it is necessary to lay out his causal theory of action in full. This is not peripheral material. It is the argument.

A. The Three Guṇas as the Engine of All Action

#

Everything in prakṛti — from the sub-atomic to the social — is some configuration of the three guṇas:

• Sattva: clarity, illumination, discrimination, harmony. A predominantly sattvic mind is more capable of viveka, more able to see clearly and choose otherwise.
• Rajas: activity, passion, drive, agitation, desire. A predominantly rajasic mind is prone to impulsive action driven by craving and aversion.
• Tamas: inertia, dullness, resistance, confusion. A predominantly tamasic mind is least capable of reflection, most swept along by conditioning, least able to interrupt the impulse-to-act cycle.

Every act — including every criminal act — is an expression of some particular mix of these three. The act of violence that emerges from a rajasic-tamasic mind is not the same kind of act as the same bodily movement performed by a sattvic mind. The guṇa-composition at the moment of action is part of the real causal story.

And crucially: no one simply chooses their guṇa-composition. It is the product of saṃskāras — the accumulated residue of past actions and experiences, themselves the product of previous guṇa-compositions. The wheel goes all the way back.

B. The Self-Perpetuating Conditioning Loop

#

Shankara’s psychology can be expressed as a loop:

Past conditioning → present guṇa-composition → present mental state → present inclination → present act → new conditioning → future guṇa-composition

At what point in this loop is there a free choice?

Shankara’s answer is viveka — discriminative insight, the capacity to see clearly what is happening and to act from that clarity rather than from the guṇa-impulse. At every node of the loop, especially in the gap between inclination and act, viveka can intervene: “this impulse arises from tamas and rajas; it is conditioned; I am not required to follow it.”

But viveka is itself a capacity. It is developed over time — over lifetimes, in Shankara’s full cosmological picture. It is not uniformly distributed. A mind dominated by tamas and rajas may have very little functional access to viveka at the moment of action. A mind developed through years of discriminative practice has more. The freedom to “have done otherwise” is, on this picture, a real but unequally distributed capacity — not a universal given that can be assumed for every adult who acts.

C. Responsibility on This Picture — Graded, Not Binary

#

Responsibility, for Shankara, is real within vyavahāra. The jīva acts. It accumulates karma. Consequences follow. Dharma and dharmic accountability are genuine features of the transactional order.

But responsibility is graded by the degree of viveka available to the actor. Someone with deeply developed discrimination had more genuine capacity to choose otherwise than someone overwhelmed by tamasic inertia at the moment of the act. This is not a licence for lawlessness. It is a more philosophically accurate picture of what “could have done otherwise” actually means.

Law, by contrast, treats capacity as binary. Either the accused had “normal capacities” or they had a recognised mental disorder. There is no intermediate space for the enormous spectrum of human freedom that Shankara’s picture implies.

The legal doctrines that do gesture toward this spectrum — diminished responsibility, provocation, youth, trauma as mitigation — are closer to Shankara’s picture than the baseline “normal adult” doctrine is. Shankara would say these partial recognitions are correct in direction but do not go nearly deep enough.

D. The Jurisprudential Consequence — Punishment Must Be Evaluated by a Specific Criterion

#

If doership is conditioned and if responsibility is graded by viveka, then punishment has a specific and demanding test it must pass. It is not “does this satisfy the victim’s family” or “does this send a message to potential offenders” or “does this express society’s condemnation.” It is this:

Does this penal practice actually reduce structural avidyā and increase viveka — or does it deepen avidyā and destroy what viveka the person had?

A prison system that reliably produces more anger, more trauma, more identification with a criminal identity, more destruction of discriminative capacity — fails Shankara’s test completely. Not because it is too harsh. Because it is doing the opposite of what punishment is for. On Shankara’s picture, punishment that increases ignorance is not punishment at all. It is a second injury, compounding the first.

This is the most concrete and actionable jurisprudential consequence in the series so far: design the penal system so that it actually reduces the avidyā-kāma-karma cycle. Examine every practice — solitary confinement, carceral violence, dehumanisation, long sentences in brutalising conditions — and ask honestly whether it increases viveka or destroys it. Then redesign accordingly.

Where Ashtavakra and Shankara Diverge on Doership

#

The divergence is real and important. It cannot be papered over.

Both accept: the real Self is akartā, pure witness. Ahaṃkāra claims “I am the doer” under ignorance. This claim is a mistake from the Self’s perspective.

Where they part:

For jurisprudence: Shankara is easier to work with. He allows a real, if conditioned, doer in the transactional world. Law can address this doer with a reasonably clear conscience, provided it calibrates its response to the degree of freedom actually available.

Ashtavakra forces the harder question: what if the entity law addresses does not exist in the way law implicitly claims? What if the “person who freely chose to offend” is a legal fiction that happens to be useful but that does not correspond to any metaphysical reality? If so, then retribution — which requires that this fiction be real — has no ultimate ground at all.

The Bhagavad Gītā Parallel — Not a Digression

#

The Gītā is Shankara’s primary scriptural source for his account of doership, and it deserves treatment as the most sophisticated philosophical model of non-doership compatible with action in any tradition.

Ashtavakra addresses those already capable of seeing through action entirely. He does not ask how to act wisely; he dissolves the question. The Gītā addresses Arjuna — who must act, who cannot escape the field of action, who must fight — and asks how one acts rightly without being the doer in the full metaphysical sense.

This is exactly the question criminal justice must eventually confront. Law cannot dissolve action. People harm each other. Consequences must follow. But on what metaphysical basis? With what understanding of agency? The Gītā’s answer — act, but do not claim ultimate authorship; be accountable, but do not mistake accountability for ultimate desert — is the closest any classical text comes to a jurisprudential model law could actually use.

Gītā 3.27: The guṇas act; the ego claims it. Already discussed in detail above.

Gītā 3.28: “Guṇas act upon guṇas.” Also discussed. This is the most compressed statement of causal action in any philosophical literature.

Gītā 4.18:

“The one who sees action in non-action and non-action in action — that one is wise among humans, accomplished in yoga, a doer of all action.”

To see action in non-action is to recognise that even the still witness is not outside the field — the field shows up in awareness. To see non-action in action is to recognise that what appears as a sovereign act is actually the guṇas moving in Prakṛti, with no ultimate self-originating author behind it. The wise person holds both simultaneously and acts from that simultaneity.

This is the model law could draw on: a conception of agency that does not require a metaphysically free first cause but that retains genuine accountability within the causal field. Conditioned accountability — responsibility that is real because it is the ego’s identification with the guṇas’ action, not because the ego stands outside the causal order and freely chose.

Gaudapada — The Question Is Ill-Formed in the Dream

#

Gaudapada’s Advaita Prakarana does not refine the account of doership. It removes the ground on which the question stands.

Mandukya Kārikā 3.39 — The Final Negation

#

na nirodho na cotpattir na baddho na ca sādhakaḥ na mumukṣur na vai mukta ity eṣā paramārthatā

Transliteration: na nirodho na cotpattir na baddho na ca sādhakaḥ / na mumukṣur na vai mukta ity eṣā paramārthatā //

Translation: “There is no dissolution, no origination, no one in bondage, no aspirant for liberation, no one seeking liberation, and no one who is liberated. This is the ultimate truth.”

If there is no origination, there is no doer — “doer” means “one from whom an act originates.” If there is no bondage, no one has been harmed in a way that calls for cosmic metaphysical redress. If there is no liberation, the whole register of “deserves freedom” and “deserves imprisonment” collapses at the highest level of analysis.

At this level, the entire penal drama — offender, victim, crime, trial, verdict, sentence — is intra-dream activity. Dream-offenders harm dream-victims in dream-societies. Dream-courts pass dream-sentences. Dream-prisons hold dream-bodies. None of this denies the felt reality of harm within the dream. It does not say suffering is not real for the one experiencing it. It denies that the response to that harm can be grounded in ultimate desert — in a cosmic moral accounting that requires the suffering of the offender to balance the suffering of the victim.

Kārikā 3.31 — Mind Generates the Field of Doer, Deed, Done-To

#

“Perceived by the mind is all this duality that we see, whether inert objects or living beings. When the mind is eventually transcended, duality is not perceived.”

If mind is the generator of the entire subject-object field, then “doer,” “deed,” and “done-to” are all mind-categories. The doer is not a metaphysical primitive — not a bedrock entity that exists prior to analysis. It is one way the mind organises its projections. When the mind stops projecting (in deep meditation, in liberation, in the cessation of the dream), neither the doer nor the deed nor the victim is found anywhere. They were products of the projection.

This is Gaudapada’s most precise contribution to the question of agency: the doer is a mind-generated organisational category applied to the appearance of action. Not real in the way a chair is real. Real in the way a dream-character is real — real within the dream, with no standing outside it.

Kārikā 3.16 — Three Levels of Student, Three Levels of Doership

#

Gaudapada explicitly distinguishes three types of student, and this stratification maps directly onto three different jurisprudential frameworks:

Inferior student (manda): Needs ritual, duties, reward and punishment language, incentives and threats. For this person, the doer is fully real. Actions have full moral weight. Law operates here. Law is correct to operate here. The entire apparatus of criminal justice — courts, police, prisons, fines — is addressed to this level.

Intermediate student (madhyama): Ready for Upanishadic reasoning and the guṇa-doership doctrine. The doer is real but conditioned. Responsibility is genuine but graded. This is Shankara’s territory. A more sophisticated criminal justice system would operate here — one that calibrates culpability to the degree of viveka available, that evaluates punishment by its effect on ignorance, that treats the offender as a conditioned being on a developmental path rather than a fixed bearer of desert.

Superior student (uttama): Ready for asparśa-yoga — “contactless yoga,” the direct seeing-through of the entire subject-object structure. The doer never existed. The deed was a mind-projection. Law has nothing to address here. This is Ashtavakra’s and Gaudapada’s domain.

The mistake jurisprudence makes is not that it operates at the first level — it must. The mistake is when it treats the first-level story as metaphysical bedrock, as a statement of ultimate truth about the nature of persons and their freedom, rather than as a pragmatically necessary convention within a particular domain of inquiry.

Kārikā 3.48 — The Final Word

#

“Never is born the essence of Jeeva… This (Jeeva) is That (Brahman) — the highest Truth where nothing is ever born.”

If the Jeeva — the individual soul, the person who appears in court, who stands accused, who serves the sentence — was never truly born in the first place, then it was never truly constituted as a doer in any ultimate sense. The “criminal” is an appearance of Brahman that has never actually separated from Brahman. The appearance acts, appears to harm, appears to deserve consequences. But what it ultimately is — Brahman — was never a doer, never a sufferer, never a criminal, never imprisoned.

This is not comfortable. But it is the consistent logical terminus of the philosophical inquiry.

The Three Traditions — Full Comparison Table

#

What Criminal Law Needs From a Doer — And What It Cannot Have

#

Criminal law’s doer is not a philosophical abstraction. It is a heavily loaded practical construct on which the entire edifice of liability rests. Law needs four things from the person it convicts:

1. Continuity of Person

#

The person in the dock today must be the same person who committed the act yesterday. Without this diachronic identity, the attribution “you did this” makes no sense. You are punishing someone for what someone else did if the “you” has changed substantially.

Challenged by all three traditions: Ashtavakra: the apparent person persists as a narrative, a story the ahaṃkāra tells, not as a metaphysical entity with fixed identity. The sakshi does not continue or discontinue. Shankara: psychological and bodily continuity is real within vyavahāra and grounds conventional responsibility. But it is continuity of a conditioned stream, not of a fixed metaphysical self. Gaudapada: personal identity over time is part of the dream narrative. There is no deep “same person” beneath the continuity of appearance. Derek Parfit arrived at the same conclusion from analytic philosophy: personal identity just is psychological continuity — there is no further deep fact, and that reductionist view undercuts strong retributivism.

2. Authorship of the Act

#

The person must have originated the act — not merely accompanied it. The act must have flowed from their decision, their intention, their free choice. Without authorship, we are not punishing a doer. We are punishing a conduit.

Challenged directly by the guṇa theory: The guṇas originated the impulse. The saṃskāras shaped the guṇa-composition. The conditioning predated the choice. The ahaṃkāra claimed authorship afterwards. At what point in this sequence did the person “author” the act in the way law requires?

3. Counterfactual Freedom

#

At the moment of action, the person could have done otherwise. They had the capacity and the fair opportunity to comply with the law. This “could have done otherwise” is what makes punishment something other than randomised suffering.

Challenged by the viveka account: The capacity to do otherwise is real — but it is viveka, discriminative awareness, and it is unequally distributed and unevenly developed. The person with deeply developed viveka had more real capacity to interrupt the guṇa-impulse than the person overwhelmed by tamas and rajas. Law’s flat “normal adult” standard misses the entire spectrum.

4. Moral Responsibility

#

The person is the appropriate target of blame, punishment, resentment, and the reactive attitudes. P.F. Strawson argued these reactive attitudes are constitutive of moral practice — they are how we relate to each other as persons, and they do not require a metaphysical foundation to be valid.

Partially accommodated: Advaita does not entirely deny this. Within vyavahāra, reactive attitudes operate and have a role. The issue is when they claim ultimate justification — when “you deserve this suffering” is presented as a cosmic truth rather than a pragmatic social claim. At that point, all three traditions say: the metaphysical foundation is not there.

Jurisprudential Implications — Each Tradition Applied to Criminal Agency

#

If Ashtavakra Is Correct

#

Mens rea tracks the intention of ahaṃkāra, which really does arise and shape behaviour in the field of appearance. That is real enough at the functional level. But law’s claim that this intention arose from a free, self-originating, uncaused choice — from a sovereign inner will standing outside the causal order — is a category error. The intention arose from conditioning, from prior saṃskāras, from the guṇa-composition of that moment. The ahaṃkāra claimed it as its own. Law took that claim at face value.

Actus reus is bodily movement affecting other bodies. This is real at the level of appearance. The causal story behind it is guṇas and conditioning. The “I” is narrative overlay.

The death penalty: It can destroy a body. It can deter some ahaṃkāras by threat of bodily destruction. It can satisfy the reactive attitudes of others. But it cannot touch the Self. The sakshi cannot be executed. The witness cannot be imprisoned. The rhetoric of ultimate payback — of finally making the person pay for what they did at the deepest level — is metaphysically false even if its social effects are real.

What survives: Strictly consequentialist justification. Punishment is justified, if at all, solely by its effects — deterrence of future harm, incapacitation of continuing dangerous patterns, rehabilitation of conditioned behaviour, protection of potential victims, social stability. Claims that “the offender deserves to suffer in their deepest self” have no ultimate ground.

If Shankara Is Correct

#

Responsibility is real within vyavahāra but it is conditioned, not libertarian. The jīva acts. The jīva accumulates karma. Dharmic consequence is real and appropriate. But no one chose to be ignorant, and the karmic cycle was never inaugurated by a free decision. Responsibility is therefore real but it operates within a field of conditioning that predates the individual’s choices.

Culpability should scale with viveka. The more developed the discriminative awareness of the actor at the time of the act, the more genuine freedom they had, and the more genuine their responsibility. This is not a binary. It is a spectrum. A deeply tamasic mind committing violence has less real freedom — less real “could have done otherwise” — than a relatively sattvic mind doing the same. Law’s uniform “normal adult” baseline is philosophically inaccurate.

Punishment must pass a specific test: Does it reduce the avidyā-kāma-karma loop? Does it increase viveka or destroy it? A prison that reliably produces more trauma, more anger, more identification with a criminal identity, more structural ignorance — fails this test completely. Not because it is too harsh. Because it deepens the very conditioning that generated the harmful act in the first place. The jurisprudential task is empirical: examine each penal practice, each sentencing regime, each prison condition — and ask honestly whether it moves the person toward greater discrimination or toward deeper ignorance. Then redesign the system accordingly.

If Gaudapada Is Correct

#

Desert has no ultimate ground. The doer was an appearance. The act was a mind-projection. The victim was also an appearance. The harm was real within the dream but not in the sense of a cosmic account that must be settled by the suffering of the offender. Retributive punishment — as the supposed settling of an ultimate moral debt — is dream-violence responding to dream-harm. It is internally coherent within the dream. It has no ground outside it.

Rights remain crucial. Within the dream, reducing suffering is the governing purpose. Rights protect dream-persons from avoidable dream-suffering. The fact that those persons are, ultimately, appearances of Brahman does nothing to reduce the felt reality of their pain within the dream. Rights are as important as ever — more important, perhaps, because they are the only protection against the misuse of “it’s all a dream” as a cover for injustice.

The most honest system: Restorative justice — focused on repairing the rupture within the community of appearances, on restoring a livable equilibrium, on preventing future harm — rather than on cosmic settling of accounts between an ultimate doer and an ultimate victim. Forward-looking, pragmatic, honest about its own basis.

The Modern Parallel — Neuroscience of Agency

#

The three Advaita positions are not alone in deflating the naive picture of doership. Contemporary neuroscience and philosophy of mind have been circling the same point from another angle, with empirical methods.

Benjamin Libet — Readiness Potentials and “Free Won’t”

#

Libet’s experiments, conducted in the 1980s and replicated since, monitored brain activity before simple voluntary movements while asking subjects to note when they became aware of deciding to move. The finding: a slow “readiness potential” — a build-up of neural activity in motor areas — begins approximately 550 milliseconds before the movement. Subjects report their conscious decision to move only about 200 milliseconds before the act.

The brain’s preparatory process is already well underway before the conscious “I decide” appears. The conscious sense of deciding is not the initiating cause of the action. It is, at minimum, a late participant in a process already begun.

Libet himself did not want this taken as a complete negation of conscious agency. He proposed “free won’t” — the conscious awareness may still have the capacity to veto, to inhibit the action in the final 200 milliseconds, even if it did not initiate it. This veto capacity maps precisely onto Shankara’s viveka: the guṇas and saṃskāras start the impulse; discriminative awareness can still interrupt it. The more developed the viveka, the more robust this veto. This is a real but limited freedom — not libertarian free will but something genuine.

Daniel Wegner — The Illusion of Conscious Will

#

In The Illusion of Conscious Will (2002), Daniel Wegner argues that the phenomenal experience of “willing” an action — the felt sense of “I am choosing this right now” — is itself a mental construction that the brain generates after the action processes have begun, as a way of making sense of what is happening.

His evidence is wide: hypnosis experiments where people perform suggested actions and claim voluntary authorship; Ouija board experiments where participants genuinely experience moving the pointer; thought-insertion phenomena in psychosis; automatic writing; cases where people claim authorship of actions they did not perform and deny authorship of actions they clearly did. The timing and apparent match between thought and action, not the hidden causal structure, determines the sense of will.

His conclusion: the experience of will is an “emotion of authorship” — a feeling the brain generates that tags certain actions as mine, as initiated by me. It does not reveal the actual causal story. It is a useful interpretive overlay, not a transparent window onto the metaphysics of agency.

This is almost exactly what the Gītā and Ashtavakra say: the guṇas act; the ego thinks “I am the doer.” The felt will is real as experience. It is not causally basic.

Patrick Haggard — The Neural Basis of the Sense of Agency

#

Haggard’s research maps the neural networks — in the supplementary motor area, the prefrontal cortex, the parietal cortex — that generate not just actions but the experience of intending and controlling those actions. Key finding: the sense of agency can be altered by transcranial magnetic stimulation, by certain drugs, by mental illness, by hypnosis. The experience of being the author of an action is a constructed neural representation, not a direct perception of metaphysical causation.

The legal implication is precise: if the sense of agency is constructed, then “I intended to do this” — the core of mens rea — is a report about a constructed representation. It may accurately track the causal process. Or it may not. Law treats it as a transparent window. Neuroscience says the window is a construction.

Strawson, Parfit — Western Philosophy’s Parallel Moves

#

P.F. Strawson argued in “Freedom and Resentment” (1962) that our practices of holding responsible are grounded not in abstract metaphysics but in our reactive attitudes — resentment, gratitude, indignation, guilt, love. These attitudes are constitutive of what it means to treat someone as a person, as a member of the moral community. Even if determinism were entirely true, these attitudes are so woven into human life that their abandonment is neither possible nor desirable.

Advaita’s response to Strawson would be nuanced: within vyavahāra, within the transactional order, these attitudes operate and have their proper place. The issue is not the attitudes themselves but the claim that they are grounded in a metaphysically ultimate fact about a metaphysically free agent. That claim is what the three traditions deny. The attitudes can survive without it — as pragmatic instruments within the transactional order, not as cosmic truths.

Derek Parfit, in Reasons and Persons (1984), argued that personal identity over time consists in psychological continuity — overlapping memories, intentions, character, and connections — with no further deep metaphysical fact behind it. On his “Relation-R” account, the future person who suffers punishment is only related by degrees of psychological continuity to the past person who offended. Strong retributivism — the claim that the very same person who freely chose to offend now deserves to suffer — loses its footing on a reductionist account of identity.

Advaita goes further in the same direction: personal identity is at best a conventional construction in mind, at worst (Gaudapada) an outright dream-projection. Parfit saw the implications himself: reductionist views about personal identity tend to weaken retributivism and strengthen more forward-looking theories of punishment.

The Convergence — What All Three Traditions and Neuroscience Agree On

#

Despite their profound disagreements about almost everything else, the three Advaita voices and the modern researchers converge on a single point that law has never absorbed:

The experience “I did this freely” is not the same thing as metaphysical free origination.

What they all agree on:

• Action is initiated before the conscious sense of “deciding” arrives (Libet’s readiness potential; Wegner’s emotion of authorship; Ashtavakra’s “ahaṃkāra claims authorship retrospectively”).
• The felt sense of willing is a constructed representation, not a transparent report from inside the causal engine (Wegner; Haggard; Gītā 3.27 — “the ego thinks ‘I am the doer’”).
• The “I” that claims authorship is not a metaphysical primitive standing outside the causal order — it is a constructed node within it (Parfit’s reductionism; Shankara’s ahaṃkāra; Gaudapada’s mind-generated doer).
• What residual freedom exists — Libet’s veto, Shankara’s viveka — is real but limited and unequally distributed, not uniformly present in every adult at every moment.

Where they disagree:

• Whether the residual freedom is real or illusory (Shankara vs Ashtavakra and Gaudapada).
• Whether responsibility within the transactional order is genuine (Shankara says yes; Gaudapada says only pragmatically).
• How much weight to give to moral practice even without a metaphysical foundation (Strawson says everything; Gaudapada says it is dream-management).

But the convergence on the central point is enough. The naive libertarian picture of the criminal — as a sharply bounded person, metaphysically outside the causal order at the moment of offence, freely initiating a harmful act from a position of sovereign inner will — does not correspond to anything that any of these inquiries, from any direction, has managed to locate.

The Open Question

#

Post 1 asked: what is the metaphysical status of the self law punishes? Post 2 asked: even if that self exists in some constructed sense, was it ever free enough to deserve punishment? Post 3 has asked: even if it had some conditioned freedom, is it really the doer of the act in the way law assumes?

Across three Advaita voices and a swathe of modern work on agency and identity, the same pattern has emerged: the “I” that law addresses as a free, self-originating, morally responsible agent is, at best, a constructed node in a causal web; at worst, a figure in a dream that never began.

That leaves a question jurisprudence can no longer avoid:

If the experience “I am the doer” is retrospective, conditioned, and constructed — if the real movers are the guṇas, prior conditioning, and the mind’s projections — what exactly is criminal law attributing responsibility to when it convicts someone?

Not: who is the appropriate target for harm-reduction and social management? That question has pragmatic answers and law answers them daily. But: who or what does law think it is punishing — and is that entity, the free, self-originating, metaphysically robust moral agent — actually there when you look?

Secondary questions to carry forward:

1. If the guṇas do the acting and the ego only claims the credit — should a genuinely just system concern itself primarily with transforming the guṇa-conditions (social, material, psychological, educational) that generate harmful acts, rather than with morally condemning the ego that claimed them?

2. If Shankara is right that viveka is the only real freedom available — should assessments of culpability be calibrated to the actor’s actual capacity for viveka at the moment of the act, rather than to a flat “reasonable person” standard that assumes uniform freedom?

3. If Wegner is right that the experience of will is constructed and Libet is right that veto-capacity is limited and variable — should law respond empirically to the fact that veto-capacity is distributed very unequally across individuals and contexts? And if so, how?

4. If Gaudapada is right that the question “who is the doer?” is ultimately ill-formed — what would an honest jurisprudence look like? One that continues to function within the dream — that addresses apparent offenders, protects apparent victims, maintains an apparent social order — while openly acknowledging that it is not settling ultimate cosmic accounts between metaphysically real doers and metaphysically real victims?

Those questions cannot be answered by importing Advaita wholesale into criminal law. They demand a slow, careful re-examination of what law is actually doing when it says “you did this” — and whether it can any longer afford to remain innocent of its own metaphysics.

Sources

#

Primary texts:

• Ashtavakra Gita — Verse 1.6 (kriyā-karma na me, manaḥ na me), Verses 1.12–1.15 (asaṅga, akriya), Verse 15.4 (na kartā na bhoktā), Chapter 18 (jīvanmukta and non-doership within action)
• Bhagavad Gita — 3.27 (guṇas act; ego thinks “I am the doer”), 3.28 (guṇas acting upon guṇas), 4.18 (action in non-action, non-action in action)
• Vivekachudamani of Adi Shankaracharya — Verse 104 (egoism as conditioned doer), Verses 254–265 (jīvanmukta and prārabdha karma)
• Mandukya Kārikā, Advaita Prakarana of Gaudapada — 3.16 (three grades of student), 3.27–3.28 (non-origination argument), 3.31 (mind generates duality), 3.39 (na nirodho na cotpattir — the full negation), 3.48 (the Jeeva was never born)

On neuroscience and agency:

• Benjamin Libet — Mind Time (2004); original experiments on readiness potentials (1983)
• Daniel Wegner — The Illusion of Conscious Will (2002)
• Patrick Haggard — research on neural correlates of the sense of agency

On personal identity and responsibility:

• P.F. Strawson — “Freedom and Resentment” (1962)
• Derek Parfit — Reasons and Persons (1984), Part III (personal identity and its implications for retributive desert)
• H.L.A. Hart — Punishment and Responsibility (1968) (capacity and fair opportunity as conditions of liability)

Next post: What is Action? — If the doer is a construction, and the guṇas do the acting, and the mind generates the field of doer-deed-done-to, what exactly is an “act”? What makes a sequence of events an act rather than mere happening? This question moves the inquiry from agency to causation — and from criminal law to the foundations of karma itself.

This post is part of the series: Thematic Study — Ashtavakra Gita, Vivekachudamani, and Mandukya Karika. The inquiry proceeds without prematurely resolving what can bear to remain open.