Computer scientist Rajesh PN Rao receives emails every week from people claiming they have cracked an ancient text that has baffled generations of scholars.

These self-proclaimed codebreakers - ranging from engineers and IT workers to retirees and tax officials - are mostly from India or of Indian origin living abroad. All of them believed they had deciphered the script of the Indus Valley Civilization, a mixture of signs and symbols.

"They claim they have solved the problem and that 'the case is closed,'" said Mr. Rao, the Huang professor at the University of Washington and the author of a peer-reviewed study of the Indus script.

MK Stalin, chief minister of the southern Indian state of Tamil Nadu, recently upped the ante, adding fuel to the fire by announcing a $1 million prize for anyone who can crack the code.

The Indus Valley Civilization, or Harappan Civilization, was one of the world's earliest urban societies, emerging 5,300 years ago in what is now northwestern India and Pakistan. Simple farmers and merchants lived in walled brick cities and prospered for centuries. Some 2,000 sites have been discovered in the area since its discovery a century ago.

The cause of the society's sudden decline is unclear, with no obvious evidence of war, famine, or natural disaster. But its greatest mystery is its undeciphered script, shrouding its language, governance and beliefs in secrecy.

For more than a century, experts including linguists, scientists, and archaeologists have been trying to decipher the Indus script. theory connects it with early BrahmiDravidian and Indo-Aryan, Sumerianseven claiming that it consists only of political or religious symbols.

However, its secrets remain locked. “Indus script is probably the most important writing system that has yet to be deciphered,” says Asko Parpola, a noted Indologist.

Today, the more popular spectacular theories equate the text with the content of Hindu scriptures and attribute spiritual and magical significance to the inscriptions.

Rao said that what most of these attempts ignored was that texts consisting of signs and symbols mostly appeared on stone seals used for trade and commerce, so they were unlikely to contain religious or mythological content.

Deciphering the Wutong script faces many challenges.

First of all, the number of characters is relatively small, about 4,000, and almost all are written on small objects such as seals, pottery, and tablets.

Furthermore, each text is brief—the average length is about five signs or symbols—with no long text on walls, stelae, or standing slabs.

Consider the common square seal: a line of markings runs along its top, an animal figure (usually a unicorn) is in the center, and next to it is an object whose meaning is still unknown.

Nothing like rosetta stonehelping scholars decode Egyptian hieroglyphs. These artifacts contain texts in both languages, providing direct comparisons between known and unknown scripts.

Recent progress in deciphering the Indus script uses computer science to solve this ancient mystery. The researchers used machine learning techniques to analyze the script, trying to identify patterns and structures that could make sense of it.

Nisha Yadav, a researcher at the Tata Institute of Fundamental Research (TIFR) in Mumbai, is one of them. Working with scientists such as Rao, her work focuses on applying statistical and computational methods to analyze undeciphered texts.

Using a digitized dataset of Indus flags from the script, they discovered interesting patterns. One caveat: "We still don't know whether the symbols are complete words, part of a word or part of a sentence," Ms. Yadav said.

Ms. Yadav and co-researchers found 67 symbols, accounting for 80 percent of the script's text. A logo that looks like a jar with two handles proved to be the most commonly used logo. Additionally, the script starts with a large number of symbols and ends with fewer symbols. Some symbol patterns occur more frequently than expected.

Additionally, a machine learning model of the script was created to recover illegible and damaged text, paving the way for further research.

“Our understanding was that the script was structured and there was an underlying logic in the writing,” Ms. Yadav said.

To be sure, some ancient scripts remain undeciphered and face similar challenges to the Indus script.

Mr. Rao quoted a script like this Proto-Elamite (Iran), Linear A (Crete), and Etruscan (Italy), whose underlying language is unknown.

Others, such as information (Easter Island) and Zapotec (Mexico) has known languages ​​"but their symbols remain unclear". The Phaistos Disk from Crete - a mysterious fired clay disk from the Minoan civilization - "closely reflects the challenge of the Indus script - its language is unknown and only one exists Known Examples".

Back in India, it's not entirely clear why Mr Stalin in Tamil Nadu has announced a reward for deciphering the script. His announcement comes on the heels of a new study linking the Indus Valley icon to graffiti found in his state.

K Rajan and R Sivananthan analyzed more than 14,000 graffiti-bearing pottery sherds from 140 excavation sites in Tamil Nadu, including more than 2,000 marks. The researchers claim that many of these symbols are very similar to the Indus script, with 60% of the symbols matching, and more than 90% of the South Indian graffiti symbols being "similar" to those of the Indus Valley Civilization.

Messrs Rajan and Sivananthan said this "suggests some cultural contact between the Indus Valley and southern India".

Many believed that Stalin's move to announce the award established him as a staunch defender of Tamil heritage and culture, pitting him against Narendra Modi's Bharatiya Janata Party (BJP), which is in power in Delhi.

But researchers believe Mr Stalin's bonus will soon go unclaimed. Scholars have compiled a complete, updated database of all artifacts known to be inscribed—vital for deciphering. "But what did the Indus people write? I wish we knew," Ms. Yadav said.