Larkana

The Indus Script: Why It Is Still Undeciphered

What the Indus script is, why it has never been read after a century of effort, and which seals and inscriptions visitors can see in Sindh and Karachi.

By Explore Pakistan · Published 25 September 2026 · 7 min read

The Indus script is the writing of the Indus Valley Civilisation, found on seals and small objects at Mohenjo-daro and dozens of other sites, and nobody can read it. There are roughly 5,000 known inscriptions and more than 400 different signs, but most inscriptions are very short and no text in a known language has ever been found beside them. So the script remains undeciphered a century after it came to light.

If you visit Mohenjo-daro, this is the writing you will see on the stone seals in the museum. This guide explains what those signs are, why they resist decipherment, where the best-known seals ended up, and how to plan a visit around seeing them in Pakistan.

At a glance
Number of inscriptionsAbout 5,000 excavated since 1924 (2025 count)
Number of signs419 in Mahadevan's 1977 list; about 694 in Wells' 2015 list
Typical lengthAround five signs
Usual directionRight to left, with exceptions
Where foundSeals, sealings, tablets, pottery, mainly across Pakistan and north-west India
Where to see them in PakistanMohenjo-daro site museum; National Museum of Pakistan, Karachi
StatusUndeciphered

What is the Indus script?

It is a set of signs, some of them pictures of things like fish, plants or people, and some of them abstract strokes. They are found mostly on small square stamp seals carved in steatite, on the clay impressions those seals left, on copper tablets and on pottery. A seal usually shows an animal, most often a bull-like creature with one horn that is often called a "unicorn", with a short line of signs above it.

The first seal with these signs to be published was a drawing by Alexander Cunningham, printed in 1875 in an Archaeological Survey of India report, before anyone knew what it belonged to. After the excavations at Harappa and Mohenjo-daro in the 1920s, it became clear that the seals came from a whole civilisation. Mohenjo-daro itself was built around 2500 BCE and abandoned around 1700 BCE, so the seals you see in a case are well over 4,000 years old.

What are the seals for?

Nobody can say for sure, though there are educated guesses. Many seals were used to stamp clay, which suggests they marked ownership or authority on goods, in trade or storage. The animal picture may have identified a person, clan or office. Some seals have a hole or boss on the back, which suggests they were worn or carried.

These are inferences from how seals were used in other ancient societies. The script itself has not confirmed them.

Why can't anyone read it?

Several problems compound one another.

  1. The inscriptions are short. Most have about five signs, and inscriptions run from one to seven lines, with single lines the most common. Reading an unknown script normally needs long texts to spot patterns.
  2. There is no bilingual text. Egyptian hieroglyphs were cracked with the help of the Rosetta Stone, which repeated one text in known languages. No such key has been found for the Indus script.
  3. The language is unknown. Nobody knows what language the Indus people spoke, so there is no way of checking a proposed reading against a spoken form.
  4. Nobody agrees how many signs there are. Iravatham Mahadevan's 1977 concordance listed 419; Bryan Wells counted about 694 in 2015. Either figure is too many for a simple alphabet and too few for a fully picture-based script, so the script is generally thought to be logo-syllabic, with signs standing for words or syllables.

Even the direction of reading is worked out indirectly. Signs are often squeezed together at the left end of a line, as if the writer ran out of room there, which suggests most lines were written right to left.

A century of attempts, in brief

YearWhat happened
1875Cunningham publishes a drawing of the first known seal
1920sExcavations at Harappa and Mohenjo-daro reveal thousands of seals
1977Mahadevan's The Indus Script: Text, Concordance and Tables lists 419 signs from 2,906 legible inscribed objects
2004Farmer, Sproat and Witzel publish "The Collapse of the Indus-Script Thesis", arguing the signs are not writing
2009Rajesh Rao, Mahadevan and colleagues argue in Science that the sign patterns resemble language
2015Wells' sign list reaches about 694 signs
2025Tamil Nadu's chief minister, M. K. Stalin, offers a US$1 million prize for a decipherment

What do scholars think it is?

Three broad positions are argued.

  • A Dravidian language. Asko Parpola, Yuri Knorozov and others have argued that the script records an early Dravidian language, and have proposed readings on that basis.
  • Not writing at all. Steve Farmer, Richard Sproat and Michael Witzel have argued that the signs stand for families, clans, gods and religious ideas, not for spoken language. Parpola replied in 2005 that their arguments "can be easily controverted".
  • An Indo-Aryan language. Proposals of this kind exist but lack scholarly consensus. One objection is that the script shows no sign of the word endings typical of Indo-European languages.

The 2009 study's own authors stressed that statistics alone do not prove the signs record language, and Sproat disputed the method. No proposal has been accepted by a majority of specialists.

Where did the famous seals end up?

Roughly 12,000 objects from Mohenjo-daro, mostly pottery sherds, were split between India and Pakistan after 1947. That division explains why some of the best-known pieces are not in Sindh.

  • The Pashupati seal, found at Mohenjo-daro in 1928 or 1929 during Ernest Mackay's excavations, shows a seated horned figure and seven signs. It measures about 3.5 cm square and is now in the National Museum of India in New Delhi.
  • The Priest-King sculpture went to Pakistan and is in the National Museum of Pakistan in Karachi, along with many stamp seals.
  • Other seals are held abroad, including a collection at the British Museum, a unicorn seal at the Metropolitan Museum of Art in New York and seals at the Musée Guimet in Paris.

Where to see Indus seals in Pakistan

At Mohenjo-daro. The site museum displays seals and replicas of seals and sealings alongside pottery, beads and figurines. The site is about 30 km by road from Larkana (a measured 30.3 km, roughly 35 to 45 minutes) and about 115 km from Sukkur. Our Mohenjo-daro museum guide covers the galleries, and Larkana to Mohenjo-daro covers the drive.

In Karachi. The National Museum of Pakistan, on Dr Ziauddin Ahmed Road in Burns Garden, has been in its present building since 1970 and has eleven galleries. Its Indus collection includes stamp seals with inscriptions and the Priest-King. We could not confirm current opening hours or the ticket price, so check before you go. Our guide to the National Museum in Karachi has more.

A few tips for looking at seals:

  • Seals are tiny, so bring a phone to zoom in on photographs if the case allows.
  • The signs on a seal are carved in reverse so that the clay impression reads the right way round. Impressions and seals are displayed side by side in many museums.
  • Read labels for "seal" and "sealing". A seal is the carved stone, and a sealing is the clay it stamped.

If you only have time for one, choose by what else you want to see. Karachi is easier to reach and holds the Priest-King; Mohenjo-daro gives you the seals in the city where they were used. For the wider story, read Mohenjo-daro history, see the Priest-King, and browse our destinations page for the rest of Pakistan.

Check the security advice first

Larkana lies in Sindh north of Nawabshah, and the UK government advises against all but essential travel to that part of the province. Pakistani visitors still go to Mohenjo-daro, but check local news and current police advice before you set out. Overseas visitors should read the UK travel advice for Pakistan, because travelling against it can void insurance.

Frequently asked questions

Has the Indus script been deciphered?

No. Despite many claims, no reading has been accepted by most specialists, and in 2025 Tamil Nadu's government offered a US$1 million prize for anyone who can do it.

How many signs are there in the Indus script?

It depends on how variants are grouped: 419 in Mahadevan's 1977 list and about 694 in Wells' 2015 list.

Which way is the Indus script read?

Most scholars think it was generally read from right to left, with some lines running left to right.

Where is the Pashupati seal?

In the National Museum of India in New Delhi. It was found at Mohenjo-daro but stayed in India when the finds were divided after 1947.

Where can I see Indus seals in Pakistan?

At the Mohenjo-daro site museum, about 30 km from Larkana, and at the National Museum of Pakistan in Karachi, which also holds the Priest-King.

Keep reading

More from Larkana

All Sindh & Balochistan guides

Timeless beauty · Serene waters · Living culture

Want us to plan it for you?

Tell us your dates, your group and where you want to go. A travel expert will shape a route for you within a day.