Seven Verses That Smell: The Rigveda's Missing Sense

The Most Chanted Verse in the Rigveda Is About a Smell

Somebody is reciting it at this moment, in a temple or a hospital room or over a phone speaker, because a person they love is ill:

tryambakaṃ yajāmahe sugandhiṃ puṣṭivardhanam

urvārukam iva bandhanān mṛtyor mukṣīya mā ‘mṛtāt

“We sacrifice to the Three Eyed One, the sweet smelling, the increaser of thriving. Like a cucumber from its stem, may I be loosed from death, not from deathlessness.”

Rigveda 7.59, verse 12. My rendering from the Saṃhitā text, checked against Griffith (1896).

This is the Mahāmṛtyuñjaya, the great death conquering mantra, the one Sanskrit verse after the Gāyatrī that has escaped its own text and gone into general circulation. It travels in the Yajurveda as well, at Taittirīya Saṃhitā 1.8.6 and Vājasaneyi Saṃhitā 3.60, and it is chanted by people who have never read a line of RV 7.59, a hymn otherwise addressed to the Maruts, at the end of which this verse to Rudra sits like a note pinned to a door. [1] [2]

Now look at the second word of it. Sugandhi, “having a good smell”. Fragrant.

Run that word family, everything built on gandha (Sanskrit: गन्ध), through the whole Rigveda and the search terminates almost immediately. Seven verses. Not seven hundred, not seventy. Seven verses out of roughly ten thousand five hundred, and the company that sugandhi keeps in that list is not devotional. There is the smell of half digested matter steaming out of a slaughtered horse’s belly. There is the smell of raw meat that the butchers are told to cook away. There is a fire that reeks of smoke, which is a fault. There is a household of women being put to sleep by a charm. There is a fire god with a sweet mouth, one compound nobody has convincingly explained, and a forest goddess who smells of ointment and is never seen.

That is the entire olfactory vocabulary of the oldest large body of poetry in any Indo-European language, a text whose descendants would go on to build one of the most elaborate cultures of scent in human history, with treatises on perfume blending, sandalwood grades, incense recipes and the aroma of gods. The Rigveda has none of it. What it has instead is seven verses, and they repay reading one by one.

Seven Verses, and the List Is Closed

Here is the complete inventory. It is short enough to print, which is itself the finding.

Two of the seven are not really about perception. Vāja-gandhya in the ninth book is opaque, and Griffith’s “smell of strength” is a translator making the best of a compound whose second member may not be the smell word at all. Sugandhi applied to Rudra is closer to an assertion of well being than a report about a scent, in the way English “sweet” runs from taste to character. Strip those and you are down to five verses in which somebody in the Rigveda actually notices an odour, and three of those five sit inside one hymn about butchering a horse.

Methods note. The counts here come from string searches over a romanized Śākala Saṃhitā, 10,528 verses in the digitized text used for this site, against the standard figure of roughly 10,550 stanzas in Aufrecht’s edition. String searching is not lemmatized: it over-collects (the name Gandharva accounts for about twenty two of the twenty nine verses containing the letters gandh, and the Gandhārī ewe of RV 1.126.7 for one more) and it can under-collect where sandhi hides a form. For exact word counts the authority is still Grassmann’s Wörterbuch zum Rig-Veda of 1873, which indexes every occurrence of every word in the Saṃhitā. [3] Treat the numbers in this article as close approximations that anyone can check, not as concordance output.

The One Passage Where the Rigveda Watches a Smell

The horse sacrifice hymn, RV 1.162, is the least abstract thing in the collection. It is a procedural poem about killing, cutting and cooking a stallion, and it names the implements: the spit, the carving boards, the sprinkling vessels, the pots. Somewhere between the slaughter and the offering, the poet becomes anxious about smell.

yad ūvadhyam udarasyāpavāti ya āmasya kraviṣo gandho asti

sukṛtā tac chamitāraḥ kṛṇvantūta medhaṃ śṛtapākaṃ pacantu

“Whatever gut content blows out of the belly, whatever smell there is of the raw flesh, let the slaughterers put that right, and let them cook the offering to a proper doneness.”

Rigveda 1.162, verse 10. My rendering, checked against Griffith (1896).

Notice the verb. Apa-vāti is from vā, “to blow”, the root behind Vāyu and vāta, the wind. The Rigveda has no verb of its own for stinking, so the smell is described the way weather is: something blows off the carcass. The remedy is heat. Cook it properly and the raw odour goes away, which is the first piece of practical olfactory reasoning in Indian literature and also the last for a very long time.

Five verses later the same worry runs the other way:

mā tvāgnir dhvanayīd dhūmagandhir mokhā bhrājanty abhi vikta jaghriḥ

“Let the fire, reeking of smoke, not make you crackle; let the glowing pot not boil over and stink.”

Rigveda 1.162, verse 15. My rendering, checked against Griffith (1896).

The second line is difficult and translators handle it differently, but the fault it fears is not in doubt: a fire or a pot that puts its own smell into the offering.

Between those two anxieties sits the one genuinely happy smell in the hymn. The crowd around the fire, waiting for meat, calls out surabhir nir hareti, roughly “it smells good, bring it out”. RV 1.162.12 is the Rigveda at its most recognizably human: people standing near a cooking fire, using their noses to decide that dinner is ready, and a poet who thought that worth putting into a liturgy.

So the text’s one sustained passage of attention to odour is not about incense or flowers or the fragrance of the divine. It is about the difference between meat that is done and meat that is not, and about smells that could ruin a ritual. For everything else in the hymn, and for the thick ritual politics around it, see the longer treatment in the horse sacrifice of RV 1.162 and 1.163; for what else was on the fire, the food of the Rigveda.

Aside. The thing that looks most like a smell word in the Rigveda is not one. Gandharva, the name of a class of divine beings who guard Soma and haunt the waters, occurs in about twenty two verses and has tempted readers since antiquity into hearing gandha in it. The etymology is not settled, the name has an Iranian relative in the Avestan Gandarəwa, and Mayrhofer’s etymological dictionary leaves it open. [4] A folk etymology that makes the Gandharvas the fragrant ones is an etymology invented by people who already lived in a world of perfume. The Rigveda does not live there.

Sugandhi, and the Cucumber

Return to the famous verse with the inventory in hand and it reads differently. Tryambaka, the Three Eyed, is Rudra, the archer god who is not yet Śiva and who is treated in the Rigveda with the caution reserved for something that can turn on you. [5] The verse asks him for release from death by way of an agricultural image so homely that it has embarrassed some translators: a ripe cucumber comes off the vine at the slightest pull, and the poet wants his death to be like that, a detachment without violence, and he wants the detachment to leave him on the deathless side.

Into that request the poet inserts sugandhi. Griffith, in 1896, did not even render it as a smell: his Tryambaka is “sweet augmenter of prosperity”, the fragrance quietly dissolved into sweetness. The word sits next to puṣṭi-vardhana, “increaser of thriving”, and the two epithets belong together. A god who smells good is a god in whose presence things flourish, milk and cattle and children and grain. Fragrance here is a symptom of abundance rather than an aesthetic experience, which is precisely how surabhi, the Rigveda’s other scent word, behaves.

Aside. The verse’s later career is a study in how a text outgrows itself. In the classical śrauta ritual it belongs to the Tryambaka offerings, cakes for Rudra presented away from the settlement, at a crossroads, in a rite whose whole logic is to keep the dangerous god at a distance from the village. [6] A verse composed to send Rudra away is now recited to invite his protection in. Nobody planned that. It is what happens to language that survives long enough.

Two Women and a Goddess

The only persons in the Rigveda described by how they smell are female, and there are two of them.

The first is a household. RV 7.55 contains a charm for putting everyone in a compound to sleep, dog included, and it works through the sleepers by where they are lying:

proṣṭhaśayā vahyeśayā nārīr yās talpaśīvarīḥ

striyo yāḥ puṇyagandhās tāḥ sarvāḥ svāpayāmasi

“The women who lie on benches, on litters, on beds; the women of auspicious scent: all of them we put to sleep.”

Rigveda 7.55, verse 8. My rendering, checked against Griffith (1896).

Puṇya-gandha is the Rigveda’s one compliment about how a person smells, and it lands in a spell. Whether it means women who have oiled themselves, or is a polite formula meaning attractive, is not recoverable; the phrase occurs once. What is recoverable is the setting, a night in an inhabited courtyard, which is the same domestic world discussed in the Rigveda’s missing architecture.

The second is not a woman but a place that has become one. Araṇyānī, Lady of the Forest, gets a single hymn, six verses long, in which the poet hears sounds in the dusk and never quite sees her. The closing verse is her portrait:

āñjanagandhiṃ surabhiṃ bahvannām akṛṣīvalām

prāham mṛgāṇāṃ mātaram araṇyānīm aśaṃsiṣam

“Smelling of ointment, fragrant, rich in food, tilling nothing: I have praised the Mother of Wild Animals, Araṇyānī.”

Rigveda 10.146, verse 6. My rendering, checked against Griffith (1896).

Four attributes, and the first two are olfactory. The one figure in the Rigveda whom the poet cannot see is the one he describes by scent, which is either a coincidence or the sharpest bit of sensory craft in the collection. Āñjana is the salve or unguent, the same word that in the Atharvaveda gets an entire hymn to itself. A goddess of the uncultivated smells like something applied to skin. For the full hymn, see Araṇyānī and the Rigveda’s one hymn deity; for the text’s broader habit of leaving things unseen, what the Rigveda hides.

Two verses is not a pattern, and I would not build a claim about Vedic gender on them. It is worth noticing all the same that the text assigns its only two scented human presences to a household of women at night and to a female wilderness, both of them at the edge of what the poet can observe.

How to Smell, in Vedic

The lexicon gets stranger when you look for the machinery of smelling rather than for smells.

The verb is ghrā. In classical Sanskrit it is ordinary. In the Rigveda it occurs once, in a hymn to Heaven and Earth, and it is not about odour:

abhijighrantī bhuvanasya nābhiṃ

“the two of them smelling at the navel of the world”

Rigveda 1.185, verse 5. My rendering. Griffith (1896) has “Kissing the centre of the world together.”

Griffith was right to write “kissing”. In Sanskrit, to smell someone is to kiss them: the affectionate gesture is an inhalation at the head, which is what a cow does to a calf and what a parent does to a child, and the verb never fully separates the two. So the Rigveda’s single use of its verb for smelling is a tenderness between the sky and the earth, bending over the point where the sacrifice happens. The sense of smell enters the text as intimacy before it enters as perception.

The nose is nearly as scarce. The noun ghrāṇa, the organ, is absent. Nāsikā occurs once, in a late healing charm that walks down a patient’s head evicting disease: “from your eyes, from your nostrils, from your ears, from your chin, from your skull, from your tongue I tear out the wasting” (RV 10.163.1). The nostril is a place where illness hides, not an instrument. The other appearance is a simile, and a lovely one, in which the Aśvins are asked to be to us what body parts are to a body:

nāseva nas tanvo rakṣitārā karṇāv iva suśrutā bhūtam asme

“Be to us protectors of the body, like the two nostrils; be for us good hearers, like the two ears.”

Rigveda 2.39, verse 6. My rendering, checked against Griffith (1896).

Even here the nostrils are guards rather than sensors, while the ears in the same breath are explicitly for hearing well. The Vedic nose is a pair of doors into the body. It is the ear that does epistemic work.

And where a modern writer would reach for smell, the Rigveda reaches for the tongue. Its verb of intimacy is rih, to lick: cows lick their calves, flames lick the wood, hymns lick at the gods. Its great sensory word is madhu, honey sweet, which saturates hundreds of verses and is discussed at length in honey and sweetness before sugar. This is a text that tastes and listens far more than it sniffs.

That gap, six hundred against twenty three, is the whole subject here. Sight and sound are not merely more frequent; they are theorized. The poets have a technical vocabulary for the inner seeing that produces a hymn, the dhī that Jan Gonda spent a book unpacking, and they are self conscious about hearing, since a mantra is by definition something heard. [7] Nothing in the Rigveda theorizes smelling, because nothing in the Rigveda does anything with a smell.

What Gets Called Fragrant

Surabhi is the workhorse, in fifteen verses, and its distribution is instructive. Sort the referents and a pattern falls out.

Almost everything on that list is either the sacrifice or the good life. Fragrance in the Rigveda is what a successful ritual and a prosperous household smell like: pressed stalks, hot ghee, cut grass, cooked meat, a well kept home. When RV 10.107.9 says the generous have won a surabhi dwelling, it is describing a reward, not a decor scheme. And the last item on the list shows how loose the word is: at RV 1.186.7 a man is surabhiṣṭama, which Griffith sensibly renders “most delightful” rather than “smelliest”. Several of the fifteen may carry no olfactory content at all.

One more absence is worth naming, because it is the kind that changes how you read the text. Agni is called dhūmaketu, “whose banner is smoke”, in six verses. Smoke is the god’s signal, the thing visible from a distance that tells you a sacrifice is underway. In a religion built around burning fat and plants, smoke is described as a sight. The single time the Rigveda mentions how smoke smells, at RV 1.162.15, it is a defect to be avoided. The gods do get the aroma of the offering in a thin handful of verses, Agni making the oblations surabhi before handing them to the Fathers at RV 10.15.12, but the Rigvedic fire is overwhelmingly a light and a heat and a mouth, and only barely a nose.

The Words That Had Not Arrived Yet

Walk into a Hindu temple today and the offering is pañcopacāra, the five services: scent, flower, incense, lamp, food. Four of those five are hard to find in the Rigveda.

Sandalwood is not in the Rigveda. Neither is aloeswood, camphor, musk, or any named aromatic. The word puṣpa appears in the plant hymn RV 10.97.15, where herbs are sorted into the flowering and the flowerless for medical purposes, which is botany rather than beauty. The one garland in the whole collection is a lotus wreath on the Aśvins at RV 10.184.2, in one of the latest and most marginal stretches of the text. There is no perfume, no perfumer, and no word for either.

Then watch how fast it fills in. The Atharvaveda, closer to the household than to the great sacrifice, devotes an entire hymn to an ointment, the āñjana of the three peaked mountain, worn as an amulet for long life. [8] In the early Upaniṣads smell becomes a category: at Bṛhadāraṇyaka 3.2.2 Yājñavalkya lists the nose among the eight graspers, each seized by its own object, and odour is the thing that grasps it. [9] By the classical schools, gandha is one of five standard sense objects and the defining quality of earth, and the analysis of how many kinds of smell exist becomes a scholastic industry that James McHugh has traced across Hindu, Buddhist and Jain sources. [10] By the first millennium of the common era there are perfumers with expertise, grades of sandalwood, temple incense, and the theological assumption that divine bodies are fragrant, the material McHugh’s Sandalwood and Carrion is built from. [11]

graph LR
    A["Rigveda: 7 gandha verses"] --> B["Atharvaveda: ointment charm AV 4.9"]
    B --> C["Upanisads: nose among the graspers"]
    C --> D["Classical schools: gandha, quality of earth"]
    D --> E["Perfumery, incense, temple offering"]
    A -.-> F["Fragrance means thriving"]
    E -.-> G["Fragrance means the divine, and a trade"]

Why a Text Can Be Nearly Smell Blind

Three explanations do most of the work, and they are not rivals.

The first is the medium. The Rigveda is not a description of Vedic life; it is the verbal instrument of a fire ritual, composed to be heard, memorized and re-performed with the accent intact. Its subject matter is skewed toward what the ritual manipulates and what poetry can do, which is why the collection can tell you a great deal about metre, heat, cattle, chariots and speech, and almost nothing about what a village smelled like at noon. A hymn is a sound object about a light source. Smell has no job in that system, and the Vedic poets, who were relentless theorists of their own craft, never gave it one.

The second is materials. The aromatics that later define Indian fragrance are southern and southeastern goods: sandalwood, aloeswood, camphor, and the resins that move along trade routes. The Rigvedic world is the Punjab and the upper Indo-Gangetic plain, and its available smells are ghee, smoke, crushed Soma stalks, wet earth after rain, cattle, cooking meat and applied unguent. Look again at the inventory above and that is exactly the list the text produces. The Rigveda is not failing to notice odours in general. It is noticing the ones it had.

The third is the most interesting, because it is not about India at all. Smell is lexically poor in most of the world’s languages. English speakers describe odours almost entirely by source, saying something smells of lemons or of burning, and perform badly at naming smells in the laboratory, which was long taken as a fact about human neurology. Then Asifa Majid and Niclas Burenhult tested speakers of Jahai, a hunter gatherer language of the Malay Peninsula that has a dozen or so abstract smell terms, and found that Jahai speakers named odours as concisely and as consistently as they named colours. [12] The inability to talk about smell is a cultural artefact, not a biological ceiling. [13] Set the Rigveda in that comparative frame and its silence stops looking like a deficiency and starts looking like a normal condition from which some cultures, India spectacularly among them, later departed.

Aside. Three cautions about arguing from absence. A missing word is not a missing experience: the people who composed these hymns had noses, oiled their skin, and could obviously tell a clean fire from a smoky one, as RV 1.162.15 proves. A corpus is not a census: what the Rigveda omits often reflects genre rather than ignorance, the same caution that applies to its thin colour spectrum in a world without blue and its odd handling of the compass in the Rigveda has no word for west. And nothing here supports a claim that Vedic speakers perceived less because they said less. The evidence licenses a statement about a lexicon and a liturgy, and stops there.

What the Nose Was For

The Rigveda is one of the most sensuously attentive documents of the ancient world, and it is attentive in two channels. It watches light with an intensity that can still stop you: the reddening of the east, the gleam of horses, the flicker on a poet’s hearth. It listens to everything, because listening is how the tradition exists at all, and it built a technology of hearing so precise that the text came down to us syllable for syllable, as described in the oral engine of Vedic transmission. Between those two channels, smell got seven verses.

And then one of those seven verses became the one that people say when they are frightened.

There is a lesson in that about how traditions actually work. The word sugandhi meant something modest in the mouth of the poet who first used it: a god associated with thriving, described the way a full milk pail or a well kept house might be described. Three thousand years later the same word is chanted in a culture where fragrance has become the primary idiom of the sacred, where the god is bathed in sandal paste and ringed with incense, and it now sounds like a description of divine presence. Nothing in the verse changed. Everything around it did.

Read them in the order the text gives them sometime: RV 1.162, which holds two of the seven, then RV 7.55, then RV 7.59, then RV 8.19, then RV 10.146, leaving aside only the one compound nobody can construe. It takes ten minutes. You will pass a slaughter, a sleeping courtyard, a plea against death, a fire with a sweet mouth, and a woman made of forest who smells of ointment and is never seen, and that is the whole olfactory world of the oldest surviving Indo-Aryan poetry. It is not much. It is astonishingly well chosen.