Walk through the ruins of Mohenjo-daro, in present-day Pakistan, and you can still see the bones of a city that knew how to live well. Covered drains ran beneath paved streets. Homes had private bathrooms. Bricks were baked to the same standard size across thousands of kilometres. This was one of the most advanced urban civilisations on Earth, flourishing alongside ancient Egypt and Mesopotamia more than 4,000 years ago.
And yet, we cannot read a single sentence its people wrote.
The Indus Valley Civilisation existed from roughly 3300 BCE to 1300 BCE, with its peak urban phase between 2600 and 1900 BCE. Its cities, mainly Harappa and Mohenjo-daro, spread across what is now Pakistan and northwest India, home to tens of thousands of people at their height. These were not isolated villages. Indus traders reached as far as the Persian Gulf and Mesopotamia, and goods linked to this civilization have turned up in excavations far from home.
What they left behind, besides drains and bricks, is writing. Thousands of small carved symbols appear on seals, pottery, copper plates, and tiny tablets, found across dozens of archaeological sites. Many of the most famous seals carry an animal, often a bull-like creature with a single horn, sitting just above a short line of symbols. Archaeologists believe these seals were pressed into clay to mark ownership of goods, much like a signature on a trade invoice today. Around 4,000 inscriptions have been recovered so far. But here is the catch: almost every single one of them is short. Most inscriptions run to about five symbols. The longest known inscription contains around 26 to 34 symbols. There is no Indus Valley equivalent of a stone slab describing a king’s victories, no scroll listing laws, no letter between two friends. Just short bursts of symbols, possibly names, titles, or trade labels, repeated across a script nobody can fully explain.
This is what makes the Indus script different from almost every other ancient writing system we have cracked. Egyptian hieroglyphs had the Rosetta Stone, a single slab carrying the same message in three scripts, including Greek, which scholars already understood. That one object unlocked an entire civilisation’s voice. The Indus script has no such shortcut. No bilingual text has ever been found. No king’s name survives anywhere else to match against a symbol. We don’t even know, for certain, what language lies underneath the signs. Over the decades, scholars have proposed early Dravidian, Sanskrit and other Indo-Aryan languages, Munda, and even a lost language related to ancient Elamite from Iran. None of these theories has been confirmed. Most have been quietly set aside when the evidence did not hold up.
This is not for lack of trying. Indologist Asko Parpola, who has spent much of his career on this puzzle, once called it perhaps the most important undeciphered writing system in the world. Researchers like Nisha Yadav at the Tata Institute of Fundamental Research in Mumbai have turned to machine learning to find patterns in how the signs are arranged. Her work found that just 67 signs account for roughly 80 percent of everything ever written in the script, and that these signs follow strict rules about where they can appear in a sequence. In other words, there is clearly a system at work, a structured logic, not random doodles. What that system actually says is still unknown.
The mystery has also turned political. In January 2025, Tamil Nadu Chief Minister M.K. Stalin announced a cash prize of one million dollars for anyone who can decipher the Indus script to the satisfaction of archaeologists. Separately, he also announced plans to establish a research chair named in honour of the late script scholar Iravatham Mahadevan. The announcements followed a study by researchers K. Rajan and R. Sivananthan, who compared thousands of pottery fragments from Tamil Nadu with Indus signs and found notable overlap between the two sets of symbols. The study’s own authors were careful to call this a comparison of shapes, not proof of a shared language. But in a country where the debate over whether ancient Dravidian and Aryan populations were separate or connected still shapes politics and identity, even a hint like this travels far.
Predictably, the million-dollar offer has invited a flood of claims. Scholars who study the script say they regularly receive messages from amateurs convinced they have cracked it. None has held up under scrutiny so far.
So the Indus Valley remains a civilisation we can map, excavate, and admire, but not quote. We know how its cities drained rainwater more efficiently than some Indian cities manage today. We do not know what its people called themselves, what gods they prayed to, or what they argued about over dinner. That gap is what keeps pulling scholars, governments, and curious amateurs back to the same handful of symbols, generation after generation. Until someone finds that one breakthrough object, an equivalent of the Rosetta Stone for the Indus world, their writing will stay exactly where it has stayed for over a hundred years: carved, visible, and silent.
Subscribe Deshwale on YouTube
