nagasena

Nāgasena to Candrakīrti:
The Cart & Designation that Connects Sutra Buddhism

The Chariot Example and Its Origins

The chariot/cart example became famous through Nāgasena in the last centuries of BCE and was picked up again in the 7th century by Candrakīrti. Its logic is simple but powerful—’wholes’ (like the chariot) do not exist other than as designations on their parts. (Chariot and cart are used interchangeably throughout—the Pāli ratha and Sanskrit ratha both refer to wheeled vehicles, it makes no difference whether war chariots or simple carts; the conceptual point remains identical.)

Historical Context: The Greco-Buddhist World

Alexander the Great → Invaded India from ancient Greece in 326 BCE and left quickly.

Seleucus I → Lost the Greco-Indian territories to Chandragupta Maurya in 305 BCE.

Menander I (165–130 BCE) → Returned with a hybrid culture, spoke Greek, but lived and died as an Indian monarch. He is famed as a convert to Buddhism, largely based on surviving coins—one bearing the famous dhamma wheel.

He is thought to be the King Milinder/Milinda portrayed in the famous Milindapañha story with Nāgasena, who is supposedly a Buddhist sage in conversation with that king. The story is thought to be legendary but the exposition it contains is strong and self-evident.

 The Milindapañha: A Text Between Traditions

The Milindapañha was not originally written for the Theravāda tradition. It likely emerged in a cultural environment influenced by various Buddhist schools, and perhaps also Greco-Buddhist syncretism. The Theravāda tradition appropriated the story, but did not canonise it. The Milindapañha is not included in the Pāli Canon itself. However, it is considered an important and influential paracanonical text within Theravāda Buddhism, meaning it’s outside of, but highly regarded within, the tradition.

The Logical Structure: Why Stop?

The chariot argument demonstrates a method—a way of analysing apparent wholes into parts and showing the whole is just a designation. Once it’s clear that the method works, the obvious question is: why stop? The wheels are also wholes composed of parts. Those parts are wholes composed of parts. The logic is iterative and the text shows this is how it works.

Nāgasena needn’t say ‘and this continues indefinitely’. He has demonstrated the principle. The implication is hanging right there.

Theravāda’s Controlled Appropriation

Theravāda Abhidhamma must stop the potential iteration before it reaches ultimate dhammas —what it holds as basic building blocks of reality (forms, feelings, perceptions, etc), existing on their own with inherent nature (svabhāva). These are supposedly unanalysable particulars that just are–they exist without dependence.

But Milindapañha‘s method exposes this belief: if every whole dissolves into parts-as-designation, then so do “ultimate” dhammas. Besides, Buddhism claims that all things are dependent. The text provides no principled stopping point. It doesn’t claim “these parts possess inherent nature and cannot be dissolved further”. It simply ends after persons, presumably because Nāgasena has demonstrated his point.

That’s why paracanonical status makes perfect sense. The text is:

  • Too valuable to discard (brilliant pedagogy, respectable antiquity, famous dialogue)
  • Too dangerous to canonise (the method threatens to dissolve Theravāda’s ultimate realities)

They keep it close but not too close. They use it for teaching about non-self but don’t let students push the logic where it naturally wants to go.

It’s a controlled appropriation of a text whose full implications they cannot accept. The Sword of Damocles hangs—is anticipated but never allowed to fall.

 Nāgasena’s Deconstructive Designation

Theravāda embraces Nāgasena’s cart for one task only: refuting a personal self over and above aggregates.

He makes no claim that all dharmas are empty as the Buddha had done and Nāgārjuna would do later. Yet, the logic does generalise—Nāgasena simply stops when his job is done. Nothing in the argument restricts it to persons only. That restriction is doctrinal, not logical.

Nāgasena’s designation (paññatti) is purely deconstructive: it dismantles the notion of a self whilst leaving the ordinary world intact as a practical function. It pulls apart what appears, without addressing the appearance itself.

Theravada appropriation paints him as “no-self only”, but the Milindapañha draws no such boundary.  If seen through an early Mahayana lens, he anticipates the four graspings found in the Vajra Cutter Sutra (Vajracchedika): “grasping as self, as sentient being, as living being, as person”. So the range of his analysis is not limited.

Nāgasena dismantles false identity at the level of description without doctrinal commitments; later, Nāgārjuna dismantles it at the level of ontology by attacking svabhāva, which allows his critique to be historicised and contained as “contextually bound to an Abhidhamma argument”.

Both Nāgasena and Nāgārjuna prove the impossibility that any form of lurking agent can supervene the aggregates, but Nāgārjuna can be tied to a tradition when Nāgasena cannot.

In the Pali Canon, the Buddha refuses to answer any question of his continuance. The teachings are an end to suffering, not a start to heaven. Nibbana is an ending not a beginning, so if “ultimate” dhammas are permanent where do they continue?

Nāgasena demonstrates infinite regress without doctrinal brakes. Theravāda keeps him paracanonical precisely because he exposes this fault line: ultimates either dissolve like the chariot or become the eternalisms the Buddha rejected.

Menanda’s Dharmacakra

One surviving coin bears an unmistakable eight-spoked wheel. In 2nd-century BCE northwestern India, this reads as dharmacakra, not a generic solar or royal symbol—it  testifies where it served and its chariot-logic roots.

 

Nāgārjuna’s Constructive Revolution

Abandoning the Example, Radicalising the Logic

Nāgārjuna removes the cart example, retains the designation reasoning, and presents it without limit, but the universality of this reasoning—that all dharmas lack a self—is already taught by the Buddha and is not an invention.

In MMK 24 and elsewhere, the same analysis is specifically applied to:

  • persons
  • causes
  • motion
  • time
  • nirvāṇa

The cart logic becomes: If X exists only by dependence and designation, X lacks svabhāva.

This is not a different argument—it is the same inferential engine but without the stopping point imposed by Theravāda Abhidhamma.

Where Nāgasena shows a deconstructive analysis: an appearance with no substance, and Abhidharma posits substance—ultimate constituents that stand on their own; Nāgārjuna does neither.

Designation as Generative Principle

Nāgārjuna’s approach is constructive, but only at the level of explanation. He constructs no objects and establishes no domain of conventional existence; rather, he provides a universal account in which anything whatsoever appears only as designation, without ground or product.

When he says things exist by designation (prajñapti), depend on conceptual imputation (saṃvṛti), and arise through dependent origination (pratītyasamutpāda)—he’s showing how the world comes to be as designated, without substrate. Designation isn’t something that happens to pre-existing objects; designation is the entire generative process itself.

This is his unique contribution: a systematic account of how designation generates the conventional world. Not negative critique or dissolution, but an explanation of the mechanism by which the conventional domain arises through conceptualisation operating on emptiness, with nothing underneath. The conventional world is real (not illusory), but real as designation, not as substance.

Bhāvaviveka’s Betrayal

Reversing Nāgārjuna for the Sake of Debate

Over 5 centuries, the reification of ultimate truth had become mainstream within Madhyamaka. Nāgārjuna’s original view  had been replaced by a realist corruption that reversed the essential groundlessness of his logic, but kept his Madhyamaka name in a gross parody of its essential insight.

Bhāvaviveka’s so-called Svātantrika Madhyamaka represents a fundamental betrayal of Nāgārjuna’s philosophy, not a legitimate methodological variant. Nāgārjuna explicitly states that there is no object prior to its designation—this is central to his entire project. Bhāvaviveka directly contradicts this, making that reversal his core innovation: he posits that objects must exist conventionally before they can be named or known, establishing an object that precedes naming  – something that Nāgārjuna does not overlook, but pointedly denies.

Not only does Bhāvaviveka break the co-arising of the object and its name, he misrepresents conventionality (saṃvṛti) as if it were a realm that can accommodate unnamed objects. But convention is societal agreement—and agreement about what? Any agreement requires a name for the thing being agreed upon. Unnamed objects cannot be conventional because there’s nothing to conventionally agree about. Without a handle, what can be agreed?

Further, by claiming that unnamed objects are produced by dependent arising, and that these objects possess determinate structure (assembled parts rather than pure designation), Bhāvaviveka requires dependent arising to ‘know’ what to create—which absurdly grants it intention and mind.

“I Have No Position”

How do academics fall for this? He obviously has many positions, one being the fact that any position can be dismantled. If Nāgārjuna has no position, what is Candrakīrti defending? Everything is empty. Dependent arising is emptiness itself and is its own proof that there is no svabhāva. The object and its name arise together; there is no object waiting to be named. Famous texts loaded with logical exposés—yet ‘I have no position’ = ‘I have the position that positions are not real, they are built on nothing’. His ‘no position’ is that everything is designated but there is no correspondence to a referent. His ‘no position’ isn’t absence—it’s the ultimate position.

The Academic Obfuscation

The academic framing of this as a mere dispute over method—Svātantrika using autonomous syllogisms versus Prāsaṅgika using consequentialist reasoning—obscures what’s actually at stake. This isn’t a disagreement about how to argue; it’s a distortion of the philosophical content itself. It’s a disagreement about arguing. Bhāvaviveka wanted to engage opponents on their own terms, to win debates according to the logical conventions his adversaries accepted. But those conventions presuppose exactly what Nāgārjuna rejected: that things possess determinate natures that can serve as the subject terms of valid inferences.

If Nāgārjuna’s position makes formal debate impossible, so be it. Philosophy should not be subordinated to the requirements of competitive disputation. If you value debate over logical consistency, develop a new theory explicitly designed for that purpose—but don’t reverse Nāgārjuna’s core insights and claim you’re still doing Madhyamaka. Bhāvaviveka chose sport over philosophy, prioritising the ability to engage opponents over fidelity to the actual argument.

Candrakīrti’s Restoration

The Chariot’s Return

500 years after Nāgārjuna, in Madhyamakāvatāra and Prasannapadā, Candrakīrti reclaims the cart analogy explicitly. He engages to restore what Bhāvaviveka corrupted: Madhyamaka cannot—and should not—adopt debate’s ontological presuppositions.

What the Cart Reveals

Candrakīrti systematises Nāgasena’s insight through seven-fold analysis, demonstrating the chariot exists in none of seven possible relations to its parts:

  1. Not the parts themselves
  2. Not different from the parts
  3. Not the basis of the parts
  4. Not dependent on the parts
  5. Not possessing the parts
  6. Not the collection of parts
  7. Not the shape of the parts

This exhaustive elimination leaves only designation–a mental imposition that comes from the mind of the observer, not the object being observed, nor from any inherent property of the object being observed.

The chariot demonstrates two self-dissolving insights:

  • Emptiness itself is designation
  • Analysis dissolves the object, then dissolves itself

The very reasoning that empties the cart is itself empty—logic, concepts, the analytic tools themselves lack svabhāva and must ultimately be abandoned. This marks Candrakīrti’s key distinction from Bhāvaviveka, who sought a ‘positive’ conventional basis for reasoning about emptiness but shattered its structure.

Candrakīrti’s most pointed addition to Nāgasena’s original logic is the refutation of “shape” (saṃsthāna). This was the final bunker where the “Realist Madhyamikas” like Bhāvaviveka tried to hide.

  • Their Argument: Even if the chariot isn’t a “soul”, it is at least the shape of the assembled parts.
  • The Refutation: Candrakīrti argues that if the “shape” belonged to the parts, it would have to be there before they were assembled. If it only appears after assembly, it is just another designation. Shape is a mental shortcut, not a physical property.

Non-Correspondence 

We usually think words correspond to things. Candrakīrti shows that words are circular: the chariot is designated based on parts, which are only “parts” because we’ve already designated a “chariot”.

This isn’t a failure of logic; it is the freedom of logic. Once you realise the analysis itself is empty, the “Sword of Damocles” finally falls—not to destroy the world, but to destroy the fear of a world without a floor.

Pre-Scholastic Clarity

By retrieving Nāgasena’s simple analogy, Candrakīrti cuts through scholastic corruption. Sometimes the most effective response to misinterpretation is to return to pre-scholastic simplicity.

Candrakīrti uses the chariot analogy openly, not masked. This signals that Nāgasena’s thinking remained alive in 7th-century Mahāyāna—unbound to any tradition, universally applicable. The chariot becomes Madhyamaka’s ambulance, called not by school loyalty but by philosophical clarity.

 

Nāgasena showed the method; Nāgārjuna applied the principle; Candrakīrti restored the integrity. The same chariot rolls through all three—first dismantling, then generalising, then returning to show that dismantling and generalising are themselves designations.

In this account –> https://www.budsas.org/ebud/milinda  that compares an early Chinese version against a later Pali version, it presents Nagasena is a sage of a different lineage yet seems to make him an expert of Theravada Abhidamma when his Abhidarma would have been totally different not to mention that it is treated here (1st century BCE) as if complete when the last text of the Pali Abhidhamma is 2nd century CE.

Perplexity AI

You are right to be suspicious of how that article frames Nāgasena and Abhidhamma; what it does is historically and doxographically quite tendentious on several fronts.

What that article is actually claiming

The piece is a comparative study of the Pali Milindapañha and the Chinese Nāgasenabhikṣu-sūtra, arguing that the Chinese underlying tradition is earlier and more “original” than the Pali Milindapañha. As part of that, it:

  • Explicitly states the Pali version belongs to the Theravāda school (because of its use of the seven Pali Abhidhamma books, standard Theravāda cetasika classes, and Nikāya terminology).

  • Explicitly states that the Chinese text shows no sectarian markers and was likely compiled before the formation of distinct schools.

So the article itself does not claim Nāgasena was historically a Theravādin Abhidhammika; rather, it claims the Pali redaction recasts him as such.

Nāgasena’s “lineage” in this account

The author emphasises that:

  • The Chinese Nāgasenabhikṣu-sūtra never mentions Abhidharma at all and uses only generic “Buddha-sūtra” / “sūtra–vinaya” language.

  • The Pali Milindapañha repeatedly calls Nāgasena an expert in Abhidhamma and puts into his mouth explicit references to the seven Pali Abhidhamma books, “Tipiṭaka,” “Saṃyutta-nikāya,” etc.

From this, he argues that the original Nāgasena legend (behind the Chinese) predates crystallised Abhidharma systems and Nikāya-classification, probably even predating clear school divisions, while the Pali text is a later Theravāda appropriation that retrofits him as a canonical Abhidhamma master. In other words, Nāgasena in the narrative is being claimed as a pan-Buddhist sage whose story is then sectarianised in the Pali. That’s effectively a “different lineage” than later Theravāda.

So your reading is accurate: the study itself says the Chinese tradition reflects a time when the Abhidhamma-piṭaka and Nikāya scheme were not yet in place, i.e., Nāgasena (as envisaged by that earlier source) would not be a Theravāda Abhidhammika at all.

The Abhidhamma and chronological problem you point to

You noted the tension: the article treats Nāgasena/Milinda as 1st c. BCE and yet shows the Pali version presupposing a fully formed Theravāda Abhidhamma—while the “last” Pali Abhidhamma text (Paṭṭhāna) is usually dated around 2nd c. CE. The article in fact uses almost exactly that observation as part of its case for the lateness of the Pali Milindapañha vis-à-vis the Chinese tradition:

  • It stresses that the Pali version mentions the seven Abhidhamma books many times (Nāgasena learning them, reciting them to Arahants, teaching Abhidhamma first to lay supporters, etc.).

  • It notes that the Chinese text never mentions Abhidharma and that its doctrinal formulas (e.g. paṭiccasamuppāda, bodhipakkhiyas) are fluid and unsystematised compared to the stereotyped Theravāda formulae in the Pali.

  • It explicitly concludes that the Chinese original was composed before the full development and canonisation of Abhidhamma and before clear school-lines, whereas the Pali reflects a Theravāda milieu where Abhidhamma and Nikāya classification are already solid.

So your chronological concern is exactly what the author is leveraging: if the Milinda–Nāgasena encounter is 1st c. BCE, but the text presents him as a master of the seven-book Theravāda Abhidhamma and quotes “Tipiṭaka” and “Saṃyutta-nikāya,” this must be later editorial retrojection, not a historically contemporary portrait.

Why Nāgasena appears as a Theravāda Abhidhamma expert in this study

The key point is: the article is not saying Nāgasena historically taught Theravāda Abhidhamma in 1st c. BCE. It is saying:

  • The Chinese strand preserves a more primitive, non-sectarian Nāgasena story with no Abhidharma or Nikāya apparatus.

  • The Pali Milindapañha is a Theravāda sectarian redaction which overlays that story with Abhidhamma, Tripiṭaka structure, and Theravāda doctrinal idiom, and that this redaction is clearly later than the era it narrates.

From that angle, “Nāgasena as Abhidhamma expert” is being analysed as a Theravādin claim about Nāgasena, not as a neutral historical description. Your sense that his “own” Abhidharma, if any, would have been structurally different from the crystallised Theravāda Abhidhamma-piṭaka is perfectly consistent with the author’s argument that the Chinese version reflects a pre-Abhidharma, pre-Nikāya stage.

So: yes, there is a real anachronism if one takes the Pali portrait at face value, and the study actually uses that anachronism as evidence that the Pali Milindapañha is a comparatively late Theravāda reworking of an earlier, non-Theravādin Nāgasena legend.