The Pillars That Sing: Hampi's Musical Stones Between Legend and Acoustics

A city of boulders

Hampi does not look like anywhere else on Earth.

On the banks of the Tungabhadra river in central Karnataka, the land is covered with enormous granite boulders, piled on top of each other in impossible-looking stacks, balanced on edges, rounded by time into shapes like giant eggs and loaves. At sunrise and sunset they glow orange and pink. Among them stand the ruins of one of the greatest cities of medieval India: temples, palaces, bazaars, aqueducts, stepwells, elephant stables and royal platforms, spread over more than forty square kilometres.

This was Vijayanagara, "the city of victory", capital of an empire that dominated southern India from the fourteenth to the sixteenth centuries. Foreign visitors who came here in its prime wrote in astonishment. The Persian ambassador Abdur Razzaq, who visited in the 1440s, said he had never seen anything to compare with it. The Portuguese traveller Domingo Paes, writing in the early sixteenth century, compared it in size to Rome.

Local tradition remembers the place as older still. The landscape around Hampi is identified with KishkindhaKishkindha The monkey kingdom of the Ramayana, where Rama met Hanuman and Sugriva., the monkey kingdom of the Ramayana, where Rama met Hanuman and Sugriva. Hills and caves around the site carry names from the epic.

In 1565, after the empire's defeat at the Battle of Talikota, the city was sacked and burned for months. It never recovered. Today Hampi is a UNESCO World Heritage Site, and its most famous monument, the stone chariot in the courtyard of the Vittala temple, appears on India's fifty-rupee note.

But the Vittala temple holds something even more remarkable than the stone chariot. It holds stone that sings.

The hall of music

The Vittala temple is dedicated to Vittala, a form of Vishnu worshipped especially in Maharashtra and Karnataka. Begun in the fifteenth century, it was greatly expanded in the early sixteenth century, during and after the reign of the empire's most celebrated ruler, Krishnadevaraya, who reigned from 1509 to 1529. It is regarded as one of the finest achievements of Vijayanagara architecture.

At its heart is a great open hall, the maha mantapa or ranga mantapa, supported by richly carved granite pillars. Many of these are not single shafts. Each main pillar is surrounded by a cluster of slender colonettescolonettes Slender secondary columns carved around a main pillar, free along most of their length., carved from the same block as the pillar, joined to it at the base and top but standing free along most of their length.

Tap one of those colonettes lightly with a knuckle or a small piece of wood, and it rings. Not a dull knock, as you would expect from stone, but a clear, sustained tone, rather like a bell or a xylophone bar. Tap another, and you hear a different note. The hall is usually said to have 56 such pillars, each with several ringing colonettes (the 2008 survey described below could find 54, four of them stripped of their musical columns), and they are often called the "SaReGaMa pillars", after the syllables of the Indian musical scale.

Tradition says that musicians and dancers performed in this hall, and that the pillars could be played as instruments during temple rituals. Whether they were played routinely, or were designed mainly as a marvel, is difficult to know. What is certain is that the effect was deliberate. Stone of this kind does not ring by accident.

Today, visitors are no longer allowed to tap the pillars. Centuries of tapping, and in recent decades enormous numbers of visitors, had begun to damage the delicate colonettes. The hall is protected, and the music is heard mainly in recordings and demonstrations.

The legend of the broken pillars

Ask a guide in Hampi about the musical pillars, and you may well hear a story.

During British rule, it is said, officials were so puzzled by the ringing pillars that they ordered two of them to be cut open, to discover what was hidden inside. Perhaps there were hollow chambers, metal rods, or some secret mechanism. When the pillars were broken, they found nothing: solid stone, all the way through. The mystery deepened, and the damaged pillars, the story goes, can still be seen.

It is a wonderful story, and it carries a real point. There is no hidden mechanism. But the story itself is hard to verify. No reliable record of such an investigation by British officials has been documented, and many pillars across Hampi were damaged during the sack of the city and in the centuries of neglect that followed. The tale is best treated as part of the site's living folklore, rather than as history.

What the story gets right is the punchline. The secret of the singing pillars was never hidden inside the stone. It is in the shape of the stone.

Why stone can sing

Every solid object vibrates when struck. The question is how it vibrates, and whether that vibration produces a tone we can hear clearly.

When you strike a slender bar of stone, metal or wood, it bends back and forth very rapidly, like a vibrating ruler held over the edge of a table. The speed of those vibrations, its frequency, determines the pitch of the note. Physicists describe this kind of vibration with what is called the theory of bending beams, developed in Europe in the eighteenth century by Leonhard Euler and Daniel Bernoulli.

The theory says that the frequency of a vibrating bar depends on a few simple things:

Its length. Longer bars vibrate more slowly and produce lower notes. In fact the frequency depends on the square of the length, so halving the length raises the pitch by four times, two full octaves.
Its thickness. Thicker bars are stiffer and vibrate faster, so they produce higher notes.
The stiffness of the material. Stiffer materials vibrate faster.
The density of the material. Heavier materials vibrate more slowly.

The pillars of the Vittala temple have actually been measured. In 2008, a team led by Anish Kumar of the Indira Gandhi Centre for Atomic Research at Kalpakkam published the first scientific study of the hall: they measured the colonettes of eleven musical pillars, recorded the sound of each one, and tested the stone with ultrasound. A few of their results:

Pillar (named for the instrument it evokes)	Ringing colonettes	Length (m)	Thickness (mm)	Lowest tones (Hz)
Saptaswara (seven notes)	8	0.96 to 1.18	76 to 100	243 to 396
Ghatam (clay-pot drum)	12	1.22 to 1.25	103 to 123	305 to 454
Panchtala (five tones)	9	0.87 to 0.95	63 to 101	442 to 588
Jaltarang (water bowls)	7	0.88 to 0.92	90 to 130	495 to 780

When the team calculated the notes that the theory of bending bars predicts for columns of those sizes, the predictions agreed well with the tones they had recorded. The ringing is the bending vibration of solid stone, nothing more mysterious, and nothing less remarkable.

So a sculptor who wants a series of different notes has two main controls: how long the free part of each colonette is, and how thick it is. Make a colonette slimmer, and its note falls. Make it shorter, and its note rises sharply.

That is the whole "secret" in principle. In practice, getting stone to sing clearly is extraordinarily difficult, for reasons we will come to.

Why it sounds like a bell, not a flute

There is one more piece of physics that explains the particular sound of the Hampi pillars.

A vibrating string, like that of a veena or a violin, produces a fundamental note together with overtonesovertones The higher tones sounding above the main note; their spacing gives an instrument its character. at exactly two, three, four and more times its frequency. Those evenly spaced overtones are what make a string's sound feel smooth and "musical".

A vibrating bar is different. Its overtones are not evenly spaced. For a bar fixed at both ends, the first overtone is about 2.76 times the fundamental frequency, and the next is about 5.4 times. Those uneven overtones give a struck bar its distinctive ringing, slightly metallic character, the sound of bells, gongs, xylophones and tuning forks rather than strings.

That is why the musical pillars of Hampi sound more like temple bells than like a flute or veena. It is also why makers of modern instruments such as the marimba carve away part of the underside of each bar, to tune its overtones into more pleasing ratios. Whether Vijayanagara sculptors attempted anything similar with the shape of their colonettes is a fascinating question for acoustic researchers.

The hardest part: the stone itself

If the physics is that simple, why don't all stone pillars ring?

Because most stone is not suitable. For a pillar to produce a clear, sustained note, the stone must be:

Uniform and fine-grained, so that vibrations travel through it evenly.
Free of cracks, voids and weathered layers, any of which absorb vibration energy and kill the ring, just as a cracked bell produces only a dull clank.
Stiff, and slow to absorb vibration, like good granite, so that a struck note rings on instead of dying away.
Shaped precisely, because small changes in length or thickness change the note.

Traditional Indian sculptors had a practical test for stone quality, and it was sound. Indian craft tradition, reflected in some of the shilpa shastra texts on sculpture and architecture, classifies stone by the sound it makes when tapped. Stone that rings clearly was considered the finest, often described as "male" stone (pum-shila) and preferred for sacred images. Stone with a duller sound was graded lower. Master sculptors in South India still tap blocks of stone and listen before they begin work. A clear ring means a sound, crack-free block. A dull sound warns of hidden flaws.

Seen in that light, the musical pillars of the Vittala temple are not a mysterious anomaly. They are the triumphant extreme of a craft that already listened to stone. Sculptors who routinely judged stone by its ring went one step further, and carved stone to ring on purpose, at chosen pitches.

The granite of the Hampi region is itself remarkable. The boulders around the site belong to some of the oldest rock formations on Earth, granites well over two billion years old, part of the ancient core of the Indian peninsula. Exposed at the surface and weathered over immense periods, they break into the rounded boulders that make Hampi's landscape so strange. The temple builders were working with excellent, ancient stone, close at hand.

The ringing rocks before the temple

There is a twist to this story that very few visitors to Hampi know.

Less than a hundred kilometres away, at a hill called Kupgal near Bellary, there is a dark, fine-grained rock formation called a dolerite dykedolerite dyke A wall of dark, fine-grained volcanic rock that forced its way up through cracks in older rock.. Its rock surfaces are covered with ancient carvings, many from the Neolithic period, thousands of years ago, first recorded by British geologists in the nineteenth century. In 2002, while studying those carvings again, the archaeologist Nicole Boivin and her colleagues noticed something extraordinary: among them were rocks that rang when struck.

These rock gongs bore the marks of countless blows, in the form of small, rounded grooves worn into the surface at the points that produced the best notes. People had come to this hill, over long periods, and played the rocks as instruments. Some researchers believe the sounds were part of rituals held at the site.

So the musical stones of the Hampi region did not begin with the Vijayanagara empire. People in this same landscape were, very probably, finding and striking ringing rocks thousands of years before the Vittala temple was built. The sculptors of Vijayanagara stood at the end of a very long local tradition of listening to stone.

Singing stones around the world

India is not the only place where people discovered that stone can make music, although few places carved it as beautifully.

Elsewhere in South India. Musical pillars are found in several other temples, including the Nellaiappar temple at Tirunelveli, the Meenakshi temple at Madurai, and the Thanumalayan temple at Suchindram, near Kanyakumari. Each has its own local traditions about how the pillars were made and played.
China. Ancient Chinese orchestras included the bianqing, a set of carefully cut L-shaped stone chimes hung from a frame. A celebrated example was found in the tomb of Marquis Yi of Zeng, buried around 433 BCE, together with a magnificent set of bronze bells.
Vietnam. In 1949, a set of long stone bars was found in the central highlands of Vietnam and identified as an ancient lithophone, a musical instrument made of stone, believed to be thousands of years old. Similar stone instruments, the đàn đá, are still played in parts of the country.
England. In the nineteenth century, a family in the Lake District built a large musical instrument from stones collected on the mountain Skiddaw, and toured with it, performing concerts.
United States. At Ringing Rocks Park in Pennsylvania, a field of boulders produces bell-like tones when struck with a hammer, a natural phenomenon that has long attracted visitors and acoustic researchers.

Everywhere, the same physics is at work. Hampi stands out not because it found a secret unknown elsewhere, but because its builders combined that physics with architecture, sculpture, music and worship in a single, extraordinary space.

The music of Vijayanagara

The singing pillars did not exist in a silent world. Vijayanagara was one of the great musical centres of Indian history, and the Vittala temple stands at the heart of that story.

The empire's rulers were generous patrons of music, dance and poetry. Foreign visitors described temple processions, festivals and royal celebrations filled with musicians and dancers. Temples maintained professional performers, and dance and music were part of daily worship.

The most important musician of that world was a saint-composer named Purandara Dasa, who lived from about 1484 to 1564. Born into a wealthy merchant family, he is said to have been a miserly moneylender until a spiritual awakening led him to give away his wealth and devote his life to music and devotion. He became a disciple of Vyasatirtha, the spiritual guide of the emperor Krishnadevaraya, and spent much of his life in Hampi.

Purandara Dasa is remembered today as the Sangita Pitamaha, the "grandfather" of Carnatic music. He is credited with organising the basic lessons that every student of Carnatic music still learns first: the graded exercises such as the sarali varisai and alankaras, taught in the raga MayamalavagowlaMayamalavagowla The raga in which every beginner in Carnatic music learns the first exercises.. If you have ever heard a child in South India practising "sa ri ga ma pa da ni sa" over and over, you have heard the method attributed to Purandara Dasa.

He signed his songs with a mudra, a poetic signature, which was the name of his god: Purandara Vittala. Vittala, the same form of Vishnu to whom the Vittala temple is dedicated. Near the river, a short walk from the temple, there is still a small pavilion known as the Purandara Dasa Mantapa, where tradition says he lived and composed, and where an annual music festival is held in his memory.

So the hall of singing pillars stood in the same city, and was dedicated to the same god, as the composer who shaped the foundations of South Indian classical music. We do not know whether Purandara Dasa ever sang beside the musical pillars, or whether they were tuned to the music of his time. But it is hard to imagine a more fitting setting for stone that sings than the temple of the god whose name was on every one of his songs.

What the claims get wrong

The musical pillars attract their share of exaggeration, and it is worth separating the common claims from what is actually known.

"The pillars are hollow, or contain metal inside." There is no evidence of this. Ultrasonic tests in 2008 found the columns to be solid granite, and the ringing is fully explained by the vibration of solid, slender stone colonettes.

"Each pillar plays a precise note of the Indian scale, tuned perfectly." The pillars produce distinct pitches, and some sets may have been shaped to give a range of notes. But the claim that they form a perfectly tuned scale has not been established by careful acoustic measurement across the whole hall, and weathering and damage over five centuries would have altered some of the notes even if they were once precisely tuned.

"The technology is a lost secret that modern science cannot explain." The physics of vibrating stone bars is well understood. What cannot easily be reproduced is the extraordinary skill of the sculptors: their knowledge of stone, their eye, their hands, and their ears. That skill was real, and much of it has indeed been lost with the decline of traditional workshops. But it is lost craft, not lost physics.

"Nowhere else in the world has singing stones." As we have seen, stone instruments and ringing rocks exist in many places. Hampi's pillars are among the most spectacular architectural examples, not the only ones.

The claims, checked

The claim	Verdict	What the evidence shows
"The pillars of the Vittala temple produce musical notes when tapped."	Confirmed	The slender granite colonettes ring with clear tones, and different colonettes produce different pitches.
"British officials cut two pillars open to find the secret."	Unverified	A popular guides' story. No reliable record of such an investigation has been documented.
"The pillars are hollow or hide a mechanism."	False	Ultrasonic tests show solid granite, and the measured tones match the bending of solid stone columns.
"The pillars form a perfectly tuned musical scale."	Not established	They produce distinct pitches, but a perfectly tuned scale across the hall has not been demonstrated, and centuries of weathering affect the notes.
"This is a lost secret science cannot explain."	False	The physics of vibrating bars is well understood. What has largely been lost is the sculptors' extraordinary craft skill.
"Ringing stones were used in this region long before the temple."	Very likely	Rock gongs at Kupgal, near Bellary, sit among Neolithic carvings, which suggests people were striking them thousands of years earlier, though the strike marks themselves cannot be dated.

Listening to stone

The pleasure of the Hampi pillars is that the truth makes them more wonderful, not less.

Imagine the scene in the early sixteenth century. A master sculptor walks among blocks of granite quarried from boulders that are more than two billion years old. He taps each one and listens, rejecting any that sound dull. He chooses the clearest, most resonant blocks. Then, with chisels and years of experience, he carves each pillar into a cluster of slender stone colonettes, judging length and thickness by eye and ear, tapping and listening and carving again until each one rings with its own voice.

He does not know the Euler-Bernoulli equation. He does not need to. His hands and ears know what the equation describes. A short distance away, in the same landscape, his distant ancestors had probably struck the ringing rocks of Kupgal thousands of years before him. He is the inheritor of an extraordinarily old human conversation with stone.

The British story of breaking open the pillars, whether or not it happened, makes a perfect parable. The investigators assumed that the secret must be hidden inside, in some mechanism. They broke the stone and found only stone. The secret had been in plain sight all along: in the shape, the material, and the knowledge of the people who carved it.

Many of the "mysteries" of ancient India are like that. We go looking for lost technologies, alien help or hidden machines, when the real marvel is in front of us: human beings, with simple tools and deep knowledge of their materials, creating things that still astonish us five hundred years later. The Hampi pillars do not need a legend to sing. They only need someone willing to listen.