A late Upaniṣadic text — the Maitrāyaṇīya Upaniṣad or Maitri Upaniṣad, conventionally dated to roughly the 2nd century BCE through the 3rd century CE in its various recensions — contains one of the earliest extant attempts to enumerate the components of yoga as a systematized practice. The text lists six components, calling them collectively ṣaḍaṅga-yoga — "six-limbed yoga." The six are: prāṇāyāma (breath-control), pratyāhāra (sense-withdrawal), dhyāna (meditation), dhāraṇā (concentration), tarka (reasoning), and samādhi (absorption).1
Look at the list. Notice what is absent. There is no āsana (postural practice). There is no yama (ethical restraint). There is no niyama (observance). When Patañjali a few centuries later formalizes yoga as aṣṭāṅga — "eight-limbed yoga" — he adds yama, niyama, and āsana, and removes tarka. The eight Patañjali limbs are: yama, niyama, āsana, prāṇāyāma, pratyāhāra, dhāraṇā, dhyāna, samādhi.
The contrast is methodologically significant. The earliest systematic enumeration of yoga's components does not include āsana. Postural practice is not part of the earliest formalized six-fold yoga. Āsana enters the systematic yoga vocabulary later — sometime between the Maitri Upaniṣad's six-fold list and Patañjali's eight-limb formalization.
This is one of the principal pieces of textual evidence for the historical claim that modern postural yoga has a shallower historical depth than its presentation suggests. Āsana as a named-component-of-yoga is post-Maitri (probably late centuries BCE through early CE). Āsana as the dominant component of yoga that it is in modern postural yoga practice is much later — probably medieval haṭha-yoga development, and dominantly modern (Krishnamacarya 1930s synthesis and successors).2
The Maitri Upaniṣad's six-limb yoga is meditation-centric in a fairly pure way. Five of the six limbs (prāṇāyāma, pratyāhāra, dhyāna, dhāraṇā, samādhi) are internal-contemplative practices. The sixth (tarka) is reasoning, which is intellectual-discursive but still internal-mental. There is no body-practice in the list at all. The six-limb yoga is purely contemplative-internal in its component-architecture.
This is consistent with White's broader polemic that Patañjala-style yoga is a history of meditation. The earliest systematic formalization of yoga as system is entirely meditative-internal. The body-practice dimension — āsana, mudrā, bandha, kriyā — enters yoga vocabulary later, and arguably as secondary additions that supplement the originally-purely-meditative core.
This is also consistent with the broader Bronkhorst-philological argument that yoga as a meditation-religious term enters Indian vocabulary in the late centuries BCE. The Maitri Upaniṣad's six-limb formalization fits this dating. The six-limb yoga is the earliest systematic articulation of the post-Bronkhorst-temporal-floor yoga-as-meditation-system.3
One of the more interesting features of the Maitri Upaniṣad list is the inclusion of tarka (reasoning) as a limb of yoga. This is unusual from a later-systematic perspective. Patañjali drops tarka and substitutes yama-niyama-āsana in the front-position of his eight-limb list. Why?
One plausible reading: the Maitri compilers were working with a formalization in which reasoning-discrimination was integral to the meditative progress. Reasoning toward the conclusion of the contemplative work was part of the work; meditation without reasoning was incomplete. The discriminative-conceptual dimension was internal to yoga, not adjacent to it.
Patañjali's formalization treats reasoning differently. Patañjali's discrimination (viveka) is a fruit of yoga-practice rather than a limb of yoga-practice. The reasoning-discrimination work happens as a result of the contemplative-disciplinary work, not as part of the contemplative-disciplinary work. The systematic-structural shift represents a re-organization of yoga's internal architecture between the Maitri and Patañjala formalizations.
The substitution of yama-niyama-āsana for tarka in Patañjali's list also represents a re-orientation — toward the ethical-disciplinary-postural foundations and away from the discriminative-reasoning foundations. Patañjali's yoga is more ethically-prepared and more bodily-prepared than the Maitri yoga, and less discursively-prepared. The differences across these two early-systematic formalizations document yoga's historical malleability even in its early systematic period.4
The Maitri Upaniṣad's absence of āsana from the six-limb list combined with the presence of āsana in Patañjali's eight-limb list lets us bracket the historical introduction of āsana as named-component-of-yoga to the interval between the two formalizations — roughly 2nd-century BCE through 2nd-century CE.
But Patañjali's āsana is not what modern yoga's āsana is. Patañjali says nothing about specific postures (no mountain-pose, no warrior-pose, no sun-salutation sequence). He says only: sthira-sukham āsanam — "the āsana is steady and comfortable" (YS 2.46). What he means is: the seat should be steady and comfortable. This is posture-for-meditation in the sense of whatever sitting-posture allows steady comfortable meditation. It is not a catalog of named postures.
The catalog of named postures — the medieval haṭha-yoga corpus's 15 to 84 to 108 named āsanas — is post-Patañjala. It emerges in the medieval period (10th-15th centuries CE) and elaborates into the modern thousands of named postures only in the post-Krishnamacarya 1930s synthesis and successors.
So āsana has three distinct historical layers:
Each layer is historically distinct. Conflating them — treating modern postural yoga as continuous with Patañjala-classical āsana and with hypothetical pre-Patañjala āsana — is the flattening operation that has dominated popular yoga-history presentations.5
You are in a yoga class. The teacher mentions that yoga has eight limbs — Patañjali's aṣṭāṅga — and lists them. You notice that āsana is one of the eight, presented as integral to yoga from the earliest systematic period. Hold the Maitri Upaniṣad historical-floor live. The earliest systematic yoga-formalization had six limbs, not eight. The six-limb yoga did not include āsana. Patañjali's eight-limb yoga is the second-known systematic formalization, and it added āsana to the earlier list.
This is not to invalidate Patañjali's eight-limb yoga or modern postural yoga. It is to historically situate them. Patañjali's yoga is one historical layer of yoga's systematic formalization; the Maitri six-limb yoga is another (earlier) layer. Modern postural yoga is much later (post-medieval haṭha-yoga through 20th-century synthesis). Each layer is real and important in its period. Treating them as continuous obscures the actual historical development.
A month later you read a yoga-history book that presents eight-limb yoga as ancient. You can hold the presentation at appropriate critical distance. The eight-limb formalization is post-Maitri and post-Bronkhorst-philological-floor. Its date is roughly 2nd century CE (Patañjali's conventional dating) — 2,000 years old, which is substantial but is not 5,000 years old, and is not the original yoga formalization.
You read a modern yoga-tradition presentation of yoga's eight limbs as the original yoga-architecture. The diagnostic: the presentation is flattening the historical development by treating Patañjali's eight-limb formalization as original. The corrective: hold the Maitri Upaniṣad six-limb formalization as the earlier available formalization. The eight-limb is Patañjali's modification of an earlier six-limb yoga, not the original yoga.
The principal tension: the Maitri Upaniṣad has multiple recensions with different dating-ranges. Some recensions may post-date Patañjali. The dating-priority of the six-limb yoga formalization relative to Patañjali is not absolutely certain. Most scholars accept the priority, but the argument is not bulletproof.
A second tension: between Maitri's six-limb list and Patañjali's eight-limb list, other yoga-component formalizations may have existed in texts that did not survive. The historical-developmental picture may be more complex than the two-formalization contrast suggests.
Open question: are there other early yoga-component formalizations preserved in partially-extant texts (Mokṣadharmaparvan of the Mahābhārata, early Pāñcarātra, early Pāśupata) that could refine the picture of yoga's systematic-historical development?
White's discussion of the Maitri six-limb yoga converges with broader scholarship on early-yoga-systematic-history (Larson and Bhattacharya on classical Sāṃkhya-Yoga, Bronkhorst on early Indian meditation, Larson on the philosophical history of Yoga as darśana). The convergence is on the historical-malleability of yoga's component-architecture across early systematic formalizations.
The pattern of earliest systematic formalization differing significantly from later canonical formalization is methodologically transferable.
History: Christian creedal history shows similar patterns. The earliest creedal formulations (1st-2nd century CE proto-creeds, the Apostles' Creed in its earliest form) differ from later canonical formulations (the Nicene Creed of 325 CE, the Athanasian Creed of 5th-6th century CE). The canonical formulations are historically-late refinements of earlier formalizations, not original creedal architecture. The cross-domain insight: canonical formulations across religious traditions frequently post-date and substantially modify the earliest systematic articulations; recovering the earliest formulations refines our historical-religious reconstruction.
Behavioral mechanics: Influence Architecture — The earliest formalizations of persuasion-and-influence systematic vocabulary (Aristotle's Rhetoric, Cicero's rhetorical works, classical sophistic teachings) differ significantly from later canonical formulations (Cialdini's six-principle system, Greene's 48-laws). The canonical formulations are modern refinements that draw on but substantially restructure earlier articulations. The cross-domain insight: systematic formulations in applied-knowledge fields routinely evolve across historical periods, and the evolutionary path is substantive, not just cosmetic.
Cross-domain (Buddhism): Buddhist meditation-component formalizations show parallel historical evolution. The early Pāli Canon's five khandhas (aggregates) and jhāna (meditative absorption) architecture differs from later Abhidharmic systematic formalizations and from much later Mahāyāna-Vajrayāna systematic formulations. The cross-domain insight: all major Indic-religious systematic formulations have historical-developmental layering, and reading them as historically flat (canonical-from-the-beginning) obscures the actual developmental complexity.
The Sharpest Implication: Modern yoga's eight-limb architecture — presented as ancient and foundational — is in fact one systematic formalization (Patañjali's, c. 2nd century CE) that modified and replaced an earlier six-limb formalization (Maitri Upaniṣad's). Modern yoga teachers and students who hold aṣṭāṅga as the yoga-architecture are working inside a historically-specific formalization that was not the original. The Maitri six-limb yoga is older and different; it included tarka (reasoning) as a limb and excluded āsana; it represents a prior possible systematic formalization that the Patañjala-tradition replaced. The implication for modern practice: the eight-limb architecture is not the only legitimate historical yoga-architecture; the six-limb alternative is also historically valid and could in principle be revived or studied as a distinct yoga tradition.
Generative Questions: