You can argue all day about where yoga "really" begins. The Upaniṣads. The Indus Valley. The Brāhmaṇas. The pre-Vedic substrate. The arguments turn on interpretation and tend to circle. Peter Schreiner did something different: he counted. He went through the Mahābhārata and tallied every appearance of the word yoga. The total: nearly nine hundred occurrences.1 Three hundred plus of them are in the Mokṣadharma Parvan alone — that long didactic section in book 12 where Bhīṣma teaches from his bed of arrows for hundreds of chapters before his own self-willed yogic death. Over a hundred more are in the Bhagavad Gītā.1
For context: in all earlier scriptural canons combined — the four Vedas, the Brāhmaṇas, the early Upaniṣads — the word yoga appears far fewer times.1 The Mahābhārata is exponential expansion in a single textual location. White uses Schreiner's count to make a structural claim: the MBh is not a place where pre-existing yoga doctrine is summarized. It is the place where the word yoga first acquires the dense textual presence that makes it a subject. The MBh is the birthplace, statistically speaking, of yoga as a Sanskrit topic with its own technical vocabulary.
The Schreiner count is a piece of philological infrastructure that anchors broader historical claims. It gives White (and the field) a quantitative reason to treat the MBh as the central textual event for the emergence of yoga as a recognized discipline. Three numbers do most of the work:
~900 total occurrences across the MBh as a whole. ~300+ in the Mokṣadharma Parvan (book 12, chapters 168–353). ~100+ in the Bhagavad Gītā (book 6, chapters 23–40).
The numbers expose where the word lived in the epic. The MDhP is by far the densest single concentration. The BhG is the culturally most prominent concentration — for modern readers, the BhG often is the MBh's yoga teaching — but the MDhP outweighs it by a factor of three. White uses this to argue that scholarly attention has been disproportionately focused on the BhG at the expense of the MDhP, which is "a richer repository of early yoga speculation than the BhG, or, for that matter, the YS or the early Upaniṣads."1
The statistical asymmetry tells you something the doctrinal analysis cannot tell you directly. The BhG is one site of yogic teaching in the epic. The MDhP is the primary site. Treating the BhG as the load-bearing text is reading the canon by cultural prominence rather than by textual density. The MDhP — Bhīṣma's hundred-chapter teaching from the arrow-bed — is the corpus that contains the operational details, the warrior-yogi narratives, the chariot metaphors, the channels-and-rays material, the yogayukta terminology, and the doctrinal architecture that the BhG only partially summarizes.
The MBh's textual history is famously layered. It accumulated across centuries, was redacted multiple times, contains both archaic Vedic material and post-Upaniṣadic philosophical material side by side. Pre-Schreiner, debates about the epic's yoga content tended to operate at the level of selected passages — this BhG verse means, this MDhP chapter argues. Schreiner's count let the conversation operate at the level of distribution.
What the distribution reveals: yoga is not uniformly distributed across the epic. Book 1 (Ādi Parvan) has relatively few occurrences. Books 2–5 have scattered references. Book 6 includes the BhG with its 100-plus mentions. Book 7 (the Droṇa Parvan) has the yogayukta warrior narratives — Bhūriśravas, Droṇa — at moderate density. Book 12 has the Mokṣadharma Parvan with its 300-plus concentration. Books 13–18 taper off again. The distribution is not random; it is concentrated at the didactic sections that were added to the epic in its later stages of compilation.
This concentration pattern tells you that the word yoga did not enter the epic with the epic. It entered with the late didactic compositions — the BhG and the MDhP — that were composed and inserted into the epic structure as the epic was being finalized. The yogic vocabulary is a late accretion. The narrative bones of the epic (the warrior story, the Pāṇḍava-Kaurava conflict, the battle) are older than the dense yoga vocabulary. The MBh is the place where pre-existing narrative material gets yoga-fied by late editors who were synthesizing the developing yogic vocabulary into a Brahminically-authorized epic frame.
This page gives the vault a quantitative anchor for the historical claim that the MBh is the textual hinge for yoga's emergence as a discipline. Whenever a vault page makes a claim about "early yoga" or "the late-Vedic / early-epic yoga context," the Schreiner statistics let you verify or qualify that the MBh — and specifically the MDhP, secondarily the BhG — is where the textual evidence actually lives.
It also gives the vault a corrective to the BhG-centrism of Anglophone yoga reception. Most English-language readers encounter yoga first through the BhG and second through the YS. The Schreiner count shows that this is a small slice of the epic-period yoga corpus. The MDhP is the larger, denser, more operationally detailed source — and it is far less widely read.
The BhG is the most translated, most commented-on, most reprinted yogic text outside the YS. Vivekānanda's Western influence accelerated this; Gandhi's twentieth-century devotion to the BhG deepened it. Modern English-language yoga education frequently teaches the BhG as the central pre-classical yogic text, with the YS as the systematizing follow-up.
The Schreiner statistics suggest this is a reception artifact rather than a textual reality. The BhG has 100-plus occurrences of yoga; the MDhP has 300-plus. If textual density were the criterion of importance, the MDhP would be the central pre-classical text and the BhG would be its abbreviated digest.
Why did the BhG win cultural prominence? Several factors. First, length: the BhG is 18 chapters and easily readable in one or two sittings. The MDhP is 186 chapters. The BhG is portable in a way the MDhP is not. Second, narrative framing: the BhG is wrapped in the dramatic encounter between Arjuna and Kṛṣṇa on the battlefield, with a clear interlocutor structure. The MDhP is Bhīṣma teaching from his arrow-bed — also dramatic, but the teaching is encyclopedic rather than dialogic. Third, theological accessibility: the BhG presents a coherent Vaiṣṇava theological frame that maps cleanly onto subsequent devotional Hinduism. The MDhP is more eclectic and harder to slot into a single sectarian frame. Fourth, modern reception: Vivekānanda, Gandhi, and twentieth-century translators amplified the BhG and largely left the MDhP alone.
The combination of factors produced the modern situation where yoga in the epic period is widely identified with the BhG and the MDhP is treated as obscure technical material. White's recovery move — using Schreiner's count to argue that the MDhP is the central pre-classical yoga text — restores the statistical reality and opens up the operational material the MDhP contains.
You have read the BhG several times. You may know parts of it nearly by heart. You have read the YS and the early Upaniṣads. You have not read the MDhP. This is the most common configuration of Anglophone yoga literacy.
Tonight you find a translation of the Mokṣadharma Parvan — Ganguli's nineteenth-century English translation is the most widely available, though Fitzgerald's recent University of Chicago Mahābhārata project is more reliable where it covers the relevant sections. You don't try to read it cover to cover. You go to one of the chapters White cites — say, Chapter 289 (the multi-body passage with Śaṅkara's quoted commentary, which becomes the foundational text for the MBh 12.289 foundational passage) — and you read just that one chapter.
What you notice first is that the yoga vocabulary is dense. It is not occasional. The chapter does not say "yoga" once and elaborate around it; it returns to yoga and yogi and yogayukta and yoga-ātmā every few verses. This is what 300 occurrences across 186 chapters looks like at the chapter level: the term is structurally present, technically used, doctrinally elaborated. You are reading the densest yoga vocabulary in the entire pre-classical Sanskrit corpus.
What you notice second is that the operational detail is finer-grained than the BhG. The BhG gives you Kṛṣṇa's teaching at the level of doctrine. The MDhP gives you Bhīṣma's teaching at the level of technique — how the yogi enters another body, what the yogayukta warrior does at the moment of self-willed death, how the channels and rays interact with the breath. This is the operational material the BhG does not contain because the BhG is structurally a summary and the MDhP is a reference work.
You read for an hour. You don't try to master the chapter. You let yourself be present to the textual density that Schreiner counted. The MDhP is not obscure technical material at the periphery of the epic; it is the yoga concentration at the center of the epic's didactic core. You read it tonight because that is where the textual evidence actually lives.
The failure mode of relying on the Schreiner statistics is treating density as the only indicator of importance. The BhG is shorter and less dense, but it is theologically integrative in a way the MDhP is not. The BhG's relatively low yoga count is partly a function of its compression — every BhG occurrence is doing more doctrinal work per word than many MDhP occurrences.
Diagnostic sign: you start treating the MDhP as definitively superior because it has more occurrences.
The honest reading uses the statistics as a corrective to BhG-centrism without inverting into MDhP-supremacism. The two texts are doing different things. The MDhP is the operational reference. The BhG is the theological synthesis. Both are central to the pre-classical yoga corpus. The statistics correct one bias; they do not establish a new one.
Schreiner's count was published as part of his work on Mahābhārata yoga vocabulary; White cites it as authoritative on the distribution of yoga occurrences across the epic.1 The count covers the critical edition of the MBh (the BORI edition, completed in the 20th century), which is the standard scholarly text. Different recensions of the MBh would yield slightly different totals, but the distributional pattern (heavy concentration in MDhP and BhG, lighter elsewhere) is robust across recensions.
The first tension is whether the yoga of the MDhP and the yoga of the BhG are the same concept or two different concepts that share a word. The doctrinal content differs. The BhG's karma-yoga / jñāna-yoga / bhakti-yoga tripartite is not the MDhP's framework. The MDhP's emphasis on yogayukta warriors and embodied ascent is not the BhG's main concern. White's reading: the word yoga in this period is in statu nascendi and "the object of a conflict of interpretations" — the dense textual presence does not mean the term has a single stable meaning.1 The 900 occurrences cover multiple competing senses of the term, and the MDhP and BhG represent different sectarian positions within an unsettled vocabulary.
The second tension is whether the MBh's yoga material reflects pre-existing oral traditions or is largely composed by the late editors. Schreiner's distribution analysis is consistent with the late-editor hypothesis (concentration in didactic sections suggests late composition), but pre-existing oral material could have been systematized by the late editors rather than composed from scratch. The Heesterman-style argument for cyclical Vedic ritual material being absorbed into yoga vocabularyHeesterman — Yāyāvara/Lāna suggests that the late editors are inheriting older substrate material and re-coding it. The 900 occurrences are partly new vocabulary applied to pre-existing material, partly inherited vocabulary that the editors are amplifying.
The third tension is what the very low yoga counts in the pre-MBh texts mean. White argues this supports Bronkhorst's "no yoga term pre-2c" thesis.2 But the few occurrences that do exist (in late Upaniṣads, in dharmaśāstra) suggest the term was in circulation before the MBh's compilation. The MBh is not inventing the word; it is amplifying the word. The amplification is what produces the 900 occurrences. Pre-MBh literature has the word but does not yet make it central. The MBh is the place where it becomes central.
White and Bronkhorst converge on reading the MBh as the textual hinge for yoga's emergence.Bronkhorst — Pre-2c No Yoga Term Bronkhorst argues that pre-second-century Indian meditation was not yet yoga in the technical sense. White uses Schreiner's statistics to anchor this in the textual record: yes, the meditation traditions are older, but the vocabulary that turns them into a discipline arrives with the MBh's compilation. The two scholars are making compatible claims at different scales.
Hopkins (1901) read the MBh's yoga material as evidence of a single coherent pre-classical yogic doctrine.MDhP Chariot of Yogic Practice Schreiner's distributional analysis modifies this: the doctrine is not single, the 900 occurrences cover competing senses of the term, and the MBh's yoga material is in motion rather than stably articulated. Hopkins's textual work remains valuable, but the broader thesis of doctrinal coherence has to be qualified by the statistical evidence of vocabulary-in-motion.
Samuel's two-period thesis adds the chronological frame.Samuel — Two-Period Thesis The MBh's compilation falls late in Samuel's substrate-period (500 BCE – 200 CE). The 900 occurrences are the textual signature of the substrate-period crystallizing into a vocabulary. The medieval Tantric and Haṭha-yogic developments are downstream of this crystallization; they inherit a vocabulary the MBh consolidated.
The statistical move — counting word-occurrences across a corpus to make distributional claims about the emergence of a concept — is methodologically transferable across domains. The Sanskrit case is one instance of a recurring analytic move.
Behavioral mechanics: Vocabulary Emergence as Discipline Marker — A discipline is not a discipline until its core terms acquire dense textual presence. Modern psychology before the 1890s had philosophy of mind and physiology; "psychology" as a discipline name acquires dense textual presence only as the practitioners begin using the word relentlessly in their own publications. The Schreiner-style count would show that "psychology" goes from rare to dense between 1879 (Wundt's lab) and 1900 (William James). The same pattern occurs in influence theory: "influence" as a stable technical term acquires density in Cialdini's 1984 book and the subsequent literature. The Sanskrit case teaches the methodological move — count where the word lives, not just what the word means — and the move applies to any discipline whose self-recognition occurs in textual record.
History: Concept Emergence Statistical Anchoring — Historians of ideas have to decide when a concept "began" being a concept. The Schreiner-style count gives a defensible answer: when textual density crosses a threshold. The MBh has 900 occurrences. The pre-MBh corpus has fewer than 50 across its entire span. The threshold-crossing happens in the MBh's compilation. This methodological move is generalizable to any history-of-ideas question (the emergence of "self" in early modern European literature; the emergence of "evolution" in biological literature; the emergence of "trauma" in psychiatric literature). The Sanskrit case is a concrete worked example of the method.
The Sharpest Implication
If the MDhP is statistically the primary pre-classical yoga text, then every Anglophone yoga curriculum that teaches the BhG and YS as the central pre-classical sources has been operating with the wrong textual canon for over a century. The implication that cuts: the operationally most detailed material in the pre-classical Sanskrit corpus has been hiding in plain sight in a text that is hard to read, long, untranslated in good English, and not packaged for modern devotional reception. Recovering the MDhP would change the textual basis of Anglophone yoga literacy substantially.
Generative Questions