Imagine a sick person in the oldest layer of Indian civilization, and two healers walking toward them. One carries a pouch of dried herbs and ground minerals. The other carries nothing in his hands at all — he carries words, charms, spells he will say over the body. Both of them are practicing ayur, healing, and both are written into the same book: the Atharva Veda. That book does not make you choose between the herb and the spell. It holds both as legitimate roads to the same cure.1
The Vedas are a collective name for four manuscripts — the Rig, Sama, Yajur, and Atharva. Dukes singles out the last one as the medical Veda. The Atharva, he writes, consists of teachings concerning two types of healing (ayur): one concerned with healing through the use of charms and spells, the other with drugs and other medicinal substances.1 Two paths, side by side. The charm-path and the drug-path are not rival schools fighting for the body; they are two columns in the same text.
The Atharva also carries its own theory of where illness comes from. It identifies sickness as arising from three causes — wind, dryness, or water (Atharva Veda 1, 12:3).2 Dukes immediately marks this as older and different from what came later: it differs, he says, from the now-standard Hindu medical classification of illness by qualities (guna), of which three are likewise identified.2 So there are two three-cause schemes stacked in history. The archaic one names physical conditions of the world — wind, dryness, water. The later one names abstract qualities, the gunas. Watching the wind-dryness-water scheme give way to the guna scheme is watching Indian medicine move from naming the weather to naming the metaphysics.
And the book itself is not a single composition. Dukes is explicit: the Atharva Veda was added to and compiled in different formats and collections by various doctors and healers over the ages.3 It is a palimpsest — layer written over layer, hand over hand, until it holds "diverse and divergent material about health." Which is exactly why he calls it an important source of Indian history.3 The contradictions inside it are not flaws; they are the strata, and you can read the history of Indian medicine in the order they were laid down.
A clean, single-author medical text tells you what one school believed at one moment. A compiled, contradictory one tells you what a whole tradition believed across centuries — if you can read the layers. That is the logic Dukes is working. The Atharva holds the charm-path and the drug-path together because, over the ages it was assembled, the tradition genuinely held both. It holds the wind-dryness-water etiology and points forward to the later guna scheme because the book outlived the first system and was still being edited when the second arrived.
So the divergence inside the text is its evidentiary value. A reader who wants doctrine wants the contradictions resolved. A reader who wants history wants them preserved — each seam marks a moment when "various doctors and healers" disagreed, and froze the disagreement into the manuscript.3 The Atharva is a core sample. Drill down through it and the strata change: spell beside drug, world-condition beside quality, archaic beside classical.
This page gives the vault its oldest Indian-medicine datum, and the rest of the Dukes medical corpus sharpens against it. The Eight Causes of Illness is the Buddhist etiology that Dukes says was "not a simple clone" of the Vedic system — and you can only measure that claim against the three-cause Atharva scheme this page records. The charm-path versus drug-path split feeds straight into Three Body Languages in Buddhist Medicine: a spell works on the mind-to-body channel, a drug on the body-to-body channel, and the Atharva's two paths are the ancestral form of that distinction. And the "compiled by various doctors over the ages" framing hands Jivaka, Doctor of the Buddha a precedent — the layered-text problem that makes a single physician hard to isolate is already visible here, in a book with no single author at all.
Take the one citation Dukes anchors hardest: Atharva Veda 1, 12:3, three causes of illness — wind, dryness, water.2 Read it against the scheme that displaced it, the guna classification with its own three.2
The archaic triad is concrete. Wind, dryness, water are conditions you can feel on your skin and see in a landscape; an illness "from wind" or "from water" sits next to the weather-causes a body actually meets. The later guna triad is abstract — qualities, not conditions. The move from one to the other is the move from a medicine that points at the world to a medicine that points at categories of being. Dukes does not editorialize the shift; he just sets the two threes side by side and notes the older is "pre-Hindu in essence."2 But the comparison does the work: you are watching a medical vocabulary climb from the elemental to the metaphysical, and the Atharva preserves the lower rung that the later tradition stepped off of.
You are dating a medical idea, and the temptation is to treat "the Vedas" as one ancient block. Lamp on, the Atharva open in front of you. First you refuse the block: Dukes says not all parts of the Vedas date from the same early time, and the Atharva especially was added to over the ages.3 So you stop asking "what does the Atharva say" and start asking "which layer." You find the three causes — wind, dryness, water — and you note that they are conditions, concrete. Then you find the later guna scheme nearby and note it is qualities, abstract. You put an arrow between them, oldest to newest, and you have a small piece of real chronology instead of a flat "ancient India believed." What survives the lamp is a layered book read as a layered book — the spell beside the drug, the wind beside the guna, each seam a date.
The source is a single passage in Dukes on Indian medicine's metaphysical basis (lines 1914–1919), citing the Atharva's two ayur-paths, the three-cause etiology at AV 1, 12:3, and the book's compiled character.123 The textual claims are firmly attributed. Dukes's framing — that the wind-dryness-water scheme is older than and distinct from the guna scheme — is his reading, plausible and consistent with the layered-text he describes, but not independently sourced here.
One flag to carry forward, in case it surfaces from elsewhere in the same book: Dukes elsewhere dates the physicians Susruta and Charaka to "around the 12th century A.D."5 That is [CONTESTED — flatly wrong]. The classical compendia attributed to Susruta and Charaka belong to the early centuries CE at the latest, with cores reaching back centuries BCE — over a thousand years before Dukes's date. The error does not touch the Atharva material on this page, but it means his medical chronology cannot be trusted on absolute dates without checking; treat his relative sequencing (archaic-before-classical) as the usable part and his absolute datings as suspect.
Open: Dukes says the Buddhist eight-cause system was "not a simple clone" of the Vedic one. Is the Buddhist eight a deliberate expansion of the Atharva's three, or an independent classification? The text asserts difference but does not trace the line.
Dukes is a practitioner-scholar reading an ancient text for a living tradition, and on the Atharva his instincts mostly serve the history well — he treats the book as a dated, layered, contradictory historical document rather than a timeless scripture, which is exactly what a critical historian like Geoffrey Samuel would want. The convergence is genuine: both Dukes and the vault's critical-historical corpus insist the Vedas are stratified, not monolithic, and that "ancient" is a range, not a point. The tension lives in Dukes's absolute datings. His Susruta/Charaka error shows a scholar whose relative chronology is careful but whose absolute pegs slip badly — the opposite of Samuel, whose whole discipline is pinning dates as tightly as the evidence allows and refusing to guess past it. Reading the two together, the rule for using Dukes on Indian medical history is clear: trust the layering, trust the sequence, check every century-number against an outside source before it enters the vault as fact.
Source Criticism of Stratified Texts — the Atharva is a small, clean instance of the general historical problem of the compiled text, and pairing them shows what the eastern-spirituality material contributes to historical method. The history page describes the technique: when a document was assembled by many hands over centuries, you read its contradictions as strata and date the layers against each other. The Atharva supplies a worked example where the strata are unusually legible — spell beside drug, world-condition beside abstract quality — because the medical content changes vocabulary as the centuries pass. The history page gives the Atharva its method; the Atharva gives the history page a case where the layers are dateable by content shift (concrete causes to guna qualities), not just by linguistic dating. The insight neither produces alone: in a medical text, you can date the seams by watching the theory of disease evolve, which is a sharper stratigraphic tool than philology alone — the science inside the scripture timestamps the scripture.
Word Versus Substance as Intervention — the Atharva's two healing paths, charm and drug, map onto the oldest split in behavioral influence: changing a person with language versus changing them with a material agent. The behavioral page studies word-as-intervention — the spell, the script, the suggestion that alters state without any substance crossing the skin. The Atharva treats that as medicine, co-equal with the pharmacological path, and never ranks them. The handshake produces a long view neither domain reaches alone: the modern placebo-and-suggestion literature is, structurally, the charm-path returning under laboratory clothes — proof that the word-intervention the ancient text filed beside drugs really does act on the body. The behavioral domain gets a deep historical root: word-as-cure is not a modern marketing trick but one of the two original branches of healing. The Atharva gets a modern vindication: the spell-path it refused to subordinate to the drug-path turns out, two and a half millennia later, to be doing measurable work.
Sharpest implication: a contradictory, many-handed text is not a worse source than a clean one — it is a better one for history, because each contradiction dates a layer, and the Atharva can be read like a core sample of Indian medicine.
Generative questions: