Evidence page · comparative
Music is older than any of its surviving instruments, and the record we have is shaped almost entirely by what does not rot. This page sets out what is actually attested, and when — from a 40,000-year-old flute in a German cave, through Mesopotamia, Egypt and the Indus, to the Sangam Tamil world — and adds one original finding: a machine count of every musical instrument named in the Rigveda. There are almost none, and the ones there are, are weapons.
c. 40,000 years ago
This is a chronology of survival, not of invention. Reed, wood, gourd, hide and string do not last. Bone, ivory and metal do. Every date above is a floor. The specialists are explicit: instruments most likely did not appear when they appear in the record, but reflect "contingent matters of raw materials available and used, vagaries of preservation, and post-depositional disturbance."
There is a second bias: the earliest flutes are in European caves partly because that is where modern humans were living in caves that preserve things. Absence elsewhere is a detection limit — the same rule this platform applies to textual silence.
the Neanderthal flute that wasn't
The "Divje Babe flute" — a juvenile cave bear femur from Slovenia, dated to about 60,000 BP and widely reported as a Neanderthal instrument — is, on the weight of subsequent analysis, almost certainly not a flute. The perforations are now generally attributed to carnivore gnawing. A minority view holds that flute status cannot be entirely ruled out.
The platform records this because it is the discipline correcting a famous and attractive claim, and because the same scrutiny is owed to every "oldest instrument" assertion, including ones about South Asia.
and the objects catalogued as lids
This is the section where the evidence has recently changed, and it is a case study in how interpretation shapes a record.
A 2025 study published in Antiquity reports a pair of third-millennium BC copper cymbals excavated in situ at Dahwa, Oman — the only well-contextualised examples from Arabia, and paralleled by Indus material and Mesopotamian iconography. The consequence for South Asia is direct: "The Dahwa cymbals, found in situ as a pair, negate an identification as lids and, by extension, support the identification of at least some of the morphologically similar objects from Mohenjo-daro and other Harappan sites as cymbals."
The same pattern appears in Mesopotamia. Similar objects from graves at Kish "were initially regarded as weapons" and only later reinterpreted as instruments, once their association with a lyre and a cylinder-seal depiction was noticed. Musical objects have twice been catalogued as something else — as containers, and as arms — until context forced a rereading.
What is now attested or strongly indicated for the Indus world:
| Evidence | Status |
|---|---|
| Copper cymbals, Mohenjo-daro and other Harappan sites | Supported by the Dahwa parallel (2025) |
| Seals depicting stringed instruments, c. 2,500 BCE | Reported in the survey literature |
| The bronze "dancing girl", Mohenjo-daro | Iconographic — implies dance, and by extension music |
| The role of music in Harappan society | "Little is known" |
And the cymbals carry a further implication. Across the wider Near East cymbals are consistently associated with ritual activity — in Mesopotamia they appear in temple-building texts of Gudea and Ur-Nammu, played by lamentation and cultic singers. The Dahwa find sits alongside the wider evidence for Indus people living and working in Oman, and the authors read it as evidence of shared musical and possibly ritual practice around the Arabian Gulf. Music travelling with traders, in other words — the same routes as the carnelian and the seals.
machine-counted, and the finding is stark
Against that background, here is an original count. Every plausible instrument term, searched across the full Rigveda and the Atharvaveda (Paippalāda recension):
| Term | Rigveda | Atharvaveda | What the contexts show |
|---|---|---|---|
| dundubhi — drum | 4 | 11 | A war drum, addressed as a weapon. RV 6.47.29–30: "drive the enemies far away… roar out… you are Indra's fist." RV 1.28.5 compares a mortar's sound to "the drum of conquerors." |
| vāṇa | 6 | 1 | At RV 1.85.10 the Maruts are dhamanto vāṇam — blowing it. A wind instrument here, not a lute. |
| godhā | 4 | 1 | Mostly the monitor lizard (RV 10.28.10). Only RV 8.69.8 is plausibly musical. |
| bakura — horn | 1 | 0 | RV 1.117.21: the Aśvins blowing the bakura against the Dasyu. A war horn, used on the enemy. |
| karkarī | 1 | 0 | RV 2.43.3 — a bird's cry compared to it. A simile, not a description. |
| śaṅkha — conch | 0 | 7 | Absent from the Rigveda entirely. |
| tūṇava / nāḍī — flute | 0 | 3 | Absent from the Rigveda entirely. |
| vīṇā | 0 | 0 | The vīṇā does not occur in the Rigveda or the Atharvaveda. In the wider Sanskrit corpus it occurs 923 times across 210 texts — all later. |
| aghāṭī — cymbal/rattle | 0 | 0 | Absent from both. |
A drum addressed as Indra's fist. A horn blown at the enemy. A wind instrument played by storm gods. A lizard. One simile. No conch, no flute, no vīṇā, no cymbal.
Set beside the archaeology, the contrast is hard to miss. The region's material record already contained metal cymbals and depictions of stringed instruments a thousand years earlier. Egypt and Mesopotamia had orchestras and written playing instructions. The Rigveda's musical vocabulary is thinner than its neighbours' and thinner than the ground beneath it — and what it does contain is oriented toward war and sacrifice rather than performance. That is a fact about a particular corpus and its concerns, not about the musical life of South Asia.
the other South Asian textual tradition
The comparison that makes the point is not with Egypt. It is with the other literature of the subcontinent. The Sangam corpus records the yāḻ with six named varieties distinguished by string arrangement and mapped onto five ecological landscapes; a professional performing class (Pāṇar, Viṟaliyar); and an indigenous fourfold classification of instruments by material — tōṟkaruvi (skin), kāṟṟukkaruvi (air), narampukkaruvi (sinew), kañcakkaruvi (bronze).
| Rigveda | Sangam Tamil | |
|---|---|---|
| Named string instruments | 0 clear | 6 named yāḻ varieties |
| Classification system | none | 4 categories, by material |
| Named performer class | none for instruments | Pāṇar, Viṟaliyar |
| Dominant context | war and sacrifice | love, war, mourning, court, dance |
Two literatures from one subcontinent, and only one of them is interested in music. That is worth stating whenever "Indian classical music" is narrated as descending from the Veda — the Vedic corpus counted here contains no vīṇā, no flute and no cymbal, and its drum is a weapon.
an honest non-answer
What the evidence supports: instrument-making in South Asia is older than any of its texts; the Indus world had metal idiophones and probably chordophones and shared musical practice with the Gulf; the Vedic corpus is not where South Asian instrumental music is documented; and the Tamil tradition preserves a developed, independently-worded system by the Sangam period.
What is not established, and is not claimed here: any line of descent from Indus instruments to later South Asian music. There is no decipherment of the Indus script, no notation, and no continuity of practice — a resemblance in the shape of a cymbal is not a musical tradition. Nor does this page claim that the Tamil system is older than the Sanskrit one, or derived from it, or that either derives from the Indus.
The honest position is that "sangītam" as a formalised system has a documented later history in the treatise literature, and a prehistory that is genuinely unknown — and that the unknown is much larger and much less Vedic than the standard narration allows.
what this page claims