A Sacred Geography of Sindh: How Sindh Built, Lost, and Rebuilt Its Holy Places

 Reading a Province by Its Shrines

Good evening. Before I show you a single map, I want you to picture something. Picture the Indus at dusk — a broad, brown, restless god of a river, sliding south through a land that receives almost no rain. Now picture, strung along its banks like beads on a thread, the domes and cupolas and whitewashed tombs of the men and women whom Sindh decided to call holy. That thread of the sacred, running from the northern gate at Kashmore down to the tidal creeks below Thatta, is the subject of today’s article. And today, we are going to read it quantitatively — as data.

The evidence in front of us is a database of 624 documented religious sites spread across all 30 districts of modern Sindh: shrines and tombs, temples and mosques, churches, gurdwaras, stupas, a Parsi fire-temple, and a synagogue. The dataset was compiled from encyclopaedic and news coverage, heritage archives, devotional websites, and district profiles. Every site is geo-located; most carry a century of foundation, a religion, a size, and a status telling us whether the place still breathes or has fallen silent.

“Islam in Sind has long been popularised and sustained by the sufi saints or pirs who fill the province’s history and whose shrines still dominate the Sindhi countryside.” — Sarah Ansari

 Let me be candid about what this is not. It is not a census. It is an archive — the residue of what could be found, verified, and pinned to a coordinate. Hundreds of sites were dropped because a date, a place, or a name could not be confirmed. So when you see a district with only a handful of dots, do not conclude it is barren; conclude that its record-keepers were fewer, or its buildings humbler, or its archives thinner. With that honest caveat in hand, the sample is still large enough — 624 places — to let the deep structure of Sindh’s sacred geography show through. Today I want to convince you of one central claim: the map of the holy in Sindh is a map of water, power, and demographic upheaval, layered one century upon the next.

 

I. The River Writes the Map

So — the promised map. Here are all 624 sites at once, each coloured by religion.  

Figure 1. Every documented site, coloured by religion. Notice how the dots hug the river corridor and thin out into the eastern desert and the western hills.

Look at the shape of it. This is not a random scatter across a rectangle. The sites crowd a north–south spine and thin dramatically to the east, where the Thar desert begins, and to the west, where the Kirthar hills rise. In an arid province, sanctity settles where life settles, and life settles where the water is. Geographers of Sindh have long insisted that the province is, quite literally, a gift of the Indus — its cultivable land, its cities, and its saints all clinging to the floodplain. Let me make the point unmissable with a satellite image.

Figure 2. Shrines on the river plain.

Why should holiness follow hydrology so faithfully? Three reasons braided together. First, population: saints preached where there were people to convert and pilgrims to gather, and people lived by the river and canals. Second, trade and route: the river was the highway, and shrines rose at fords, ferries, and river-ports — Rohri, Sukkur, Sehwan, Thatta — where travellers rested, and blessings were bought. Third, and most hauntingly, the river moves. The Indus has repeatedly changed its bed over the centuries, and when it abandoned a channel it often abandoned a town — and the shrines with it. The desiccation of old courses is one quiet reason why some medieval sites in the database now stand as ruins in the middle of nowhere, stranded by a river that walked away.

 

II. The Great Acceleration — Why Shrines Explode After 1700

Now let us pull the same 624 sites apart along a second axis: time. When were they built? The answer is dramatic, and it is the single most important pattern in the whole dataset.

Figure 3. Foundations by century, stacked by religion. Fewer than 50 documented sites predate the 16th century; over three-quarters were founded from the 18th century onward.

Hold that image. From the 5th to the 15th century — a full millennium — the record gives us a thin trickle: Buddhist stupas, a scattering of early Muslim tombs, the first Hindu and Jain sites. Then, from the 16th century, the curve bends upward, and from the 18th century it detonates. Roughly three out of every four sites in the database were founded in or after the 1700s. Cumulatively, the compounding is even starker:

Figure 4. Cumulative foundations. The sacred landscape does not grow steadily — it hockey-sticks in the last three centuries.

So what happened in the eighteenth century?

Three forces, arriving almost on top of one another. The first was indigenous Muslim state-building. The Kalhora dynasty rose as the Mughal empire weakened after 1707, achieving sovereignty by the 1730s and founding the first indigenous Muslim state in Sindh. The Kalhoras and, after 1783, the Talpurs were prodigious builders of tombs and mosques; scholars of Kalhora architecture note that their output was overwhelmingly funerary and devotional — mausolea, shrines, and mosques rather than palaces. A dynasty that legitimised itself through saintly descent naturally left the landscape thick with domes.

The second force was the maturing of the pir institution. Sarah Ansari’s classic study shows that by the 18th and 19th centuries the hereditary sajjada nashins — the men who sat on a saint’s ‘cushion’ and inherited his spiritual authority — had become a landed religious elite whose shrines anchored both faith and rural power. Every generation of a pir lineage could add a tomb; every successful shrine spawned satellite shrines of disciples. The institution was, in effect, a self-replicating machine for producing sacred sites.

The third force was colonial modernity. After the British annexed Sindh in 1843, railways, canal colonies, and the explosive growth of Karachi as a port created whole new congregations — and whole new kinds of sacred building. That is why the 20th-century bar in Figure 3 is not just taller but more colourful: it is the moment when churches, gurdwaras, and formalised temples join the Muslim shrines in large numbers.

A necessary word of caution, because we are being honest with our data: some of this steepening is a survivorship effect. Recent buildings are simply likelier to survive, to be photographed, and to be dated with confidence than a 12th-century structure long since dissolved into a mound. The true medieval landscape was richer than our 49 pre-1500 dots suggest. But survivorship cannot explain the whole surge, because we have independent historical testimony — Kalhora building programmes, the rise of the pirs, the colonial city — that genuinely multiplied sacred sites. The curve is part artefact, mostly real.

 

III. A Muslim Majority in a Stubbornly Plural Land

Let us now ask the communal question directly. Whose sites are these?

Figure 5. Composition of the database by religion.

About seven in ten documented sites are Muslim — overwhelmingly the shrines and tombs of Sufi saints and pirs. But notice what the other three in ten tell us. Hinduism, Christianity, Sikhism, Jainism, Buddhism, Zoroastrianism, and Judaism are all present. This is the statistical signature of a truth every historian of Sindh repeats: the province was, and to a surprising degree remains, plural. Even today, non-Muslims make up close to a tenth of Sindh’s population, comparable to neighbouring Indian states across the border.

What kinds of building embody this sacred order? Overwhelmingly, the tomb and the shrine.

Figure 6. Sites by architectural type. The shrine–tomb complex dominates; it is the defining sacred form of Sindh.

Shrines and tombs together account for well over half of everything in the database. This is the material expression of a particular kind of Islam — one built not around the congregational mosque alone but around the grave of the friend of God. Anthropologists describe the Sindhi shrine as a space where the boundary between communities has historically been porous: a place where, in the words of one ethnographer, both Muslims and non-Muslims “take part in happiness activities together”. The classic emblem is Odero Lal, a single-walled complex where Muslims venerate Sheikh Tahir on one side, while Hindus perform aarti to the river-saint Jhulelal on the other — the same sanctity, split by a wall, shared across a faith line.

And the four largest communities do not sit in the same places. Watch how their footprints differ.

Figure 7. The same province, one faith at a time. Muslim sites blanket the whole valley; Hindu sites lean toward towns and the eastern desert districts; Christian sites concentrate sharply in colonial urban centres; Sikh sites cluster in the old commercial towns of upper Sindh.

These are not just pretty pictures; they are social history. The Muslim blanket reflects centuries of rural conversion under the pirs. The Christian sites huddle in Karachi and the cantonment towns because they are a colonial and urban phenomenon — missions, garrisons, and railway congregations. The Sikh gurdwaras trace the old Hindu-Sikh trading diaspora of Shikarpur and Sukkur, upper Sindh’s mercantile heartland. And Hindu sites reach furthest east, into Tharparkar and Umerkot, the desert districts that remained most heavily Hindu through Partition and beyond.

 

IV. Silence and Survival — What the Status of a Site Reveals

Here is where the database becomes genuinely revealing about power, because it records not just when a site was born but whether it is still alive. Let us compare survival across communities.

Figure 8. Status by religion, as a share of each community’s sites. Read the coloured bands as fates: active/heritage, damaged, disused, ruined, lost.

Two stories jump out. The first is that Buddhism and Jainism survive only as ruins. Every Buddhist site in the database — a handful of ancient stupas — is a ruin, and most Jain temples likewise. These faiths were once vital in Sindh: the Chinese pilgrim Xuanzang, visiting a half-century before the Arab conquest, described a landscape full of Buddhist monasteries. After 712 CE the region’s Buddhists gradually adopted Islam and Buddhism faded, leaving only mounds. Here the database is a time-capsule of a religious extinction that took centuries.

The second story is subtler and concerns Islam itself. A striking number of Muslim sites are marked not active — chiefly small, old tombs whose saint has been forgotten, whose lineage died out, or whose village the river left behind. So even within the dominant tradition, the sacred landscape is constantly composting: minor shrines fall silent as major ones swell. Sanctity in Sindh is not a fixed stock but a flow — sites are perpetually founded, promoted, neglected, and abandoned.

Figure 9. Among inactive, ruined, and lost sites, tombs and temples dominate — the two most vulnerable forms.

But notice what the status data does not show. Given the trauma of Partition, you might expect the Hindu sites in the database to be overwhelmingly ruined or lost. They are not: a large majority of documented Hindu sites are recorded as active or heritage-listed. This is partly real — Sindh’s temples, unlike Punjab’s, remained comparatively active because so many Sindhi Hindus, especially in the desert districts, never left. But it is partly a selection effect in our archive: a temple that survives is a temple that gets photographed, coordinated, and entered into a database. The hundreds that were demolished, encroached, or quietly converted after 1947 are precisely the ones that left no verifiable trace — and so were dropped from the dataset. The silence in our data is itself a historical fact.

 

V. 1947 and the Rewriting of the Sacred Map

No honest reading of Sindh’s sacred geography can avoid the rupture of Partition. So let us confront it with the data. Recall Figure 3’s tall, multicoloured 20th-century bar. Zoom into the minority faiths within it:

Figure 10. The colonial-era bloom of minority sites. Christian and Sikh foundations are almost entirely a 19th–20th century, urban, colonial phenomenon.

Every documented Christian site in Sindh was founded in the 19th or 20th century, and the large majority of Sikh sites are 20th-century. These communities’ sacred buildings are children of the colonial city — of Karachi’s docks, the cantonments, the railway towns. This sets up the cruel irony of 1947. Sindh, alone among the great partitioned provinces, saw almost no Partition massacre; violence here was ‘quite restrained’ compared with the Punjab. And yet the demographic transformation was total. When communal violence finally reached Karachi in January 1948, the province’s urban Hindus — who had formed the commercial majority of cities like Hyderabad, Shikarpur, and Sukkur — sailed en masse for Bombay. Sindh fell from over a quarter Hindu in 1941 to under a tenth by 1951.

What happened to the buildings they left? They passed to a state body — the Evacuee Trust Property Board — which was meant to be their guardian but too often “let them crumble or lost them to encroachment”. In the heritage and news archives, you can see the aftershocks directly: the notes on individual sites record temples “occupied by Evacuee Trust,” land “sold by a parliamentarian,” a shrine “closed in 1992 after the Babri Mosque incident, reopened 1999,” others “attacked in 1992,” “desecrated in 1993,” “attacked in 2014.” Each of these is a datapoint where national and even international politics reached down and touched a single sacred stone. The 1992 demolition of the Babri Masjid in India rippled straight back into Sindh, where mobs attacked temples across the province in retaliation.

So the sacred map of Sindh is not a serene inheritance. It is a political sediment. Sites appear when a community rises — the Kalhora tombs, the colonial churches. Sites fall silent or vanish when a community is displaced — the Buddhist stupas after the 8th century, the urban temples after 1947. The presence or absence of a dot on our map is, very often, a fossil of who held power in that place, and who had to leave.

 

VI. The City and the Countryside

Two final lenses. First, the districts — where does the record thicken?

Figure 11. Documented sites per district, shaded by the share that are non-Muslim. Karachi, the old river-towns, and the deep-Hindu desert districts stand out.

The leaders are telling: South Karachi, Thatta, Hyderabad, Sukkur — the great river-ports and the colonial-modern metropolis. Thatta’s prominence is no surprise to any historian: its Makli necropolis is one of the largest funerary landscapes on earth, the burial ground of a medieval capital. The colouring adds a second dimension: the districts richest in non-Muslim sites are the desert east (Tharparkar, Umerkot, Mirpur Khas) and the old mercantile towns — exactly where Sindh’s Hindu, Jain, and Sikh trading communities were densest. Now step into the city itself.

Figure 12. Greater Karachi in close-up. In the space of a single metropolis you find Muslim shrines, Hindu temples, churches, a gurdwara, and a Parsi fire-temple within a few kilometres of one another.

This little constellation is the whole argument in miniature. Karachi grew from a fortified fishing village into one of Asia’s great ports, and as it grew it drew in Hindu and Parsi merchants, Goan and Tamil Christians, Sikh traders, Baghdadi Jews, and Muslim migrants from across the subcontinent. Each community printed its sanctity onto the same few square kilometres. The result — visible right there in the scatter — is the densest, most plural sacred grid in the province.

 

VII. A Note on Scale — Grand Shrines and Humble Graves

One last chart, because it corrects a romantic error. When we say ‘shrine’ we tend to imagine the great illuminated darbars — Lal Shahbaz Qalandar at Sehwan, Shah Abdul Latif at Bhit Shah, Sachal Sarmast. But most sacred sites in Sindh are small.

Figure 13. Size by type. The famous ‘large’ shrines are the exception; the landscape is built overwhelmingly from small and medium tombs and mosques.

This matters for how we understand devotion. The sacred landscape is not a few cathedrals in a void; it is a dense capillary network of local, humble, often anonymous graves — a saint for every few villages, a tomb at every crossroads. The grand shrines are simply the nodes where this capillary system pools into something monumental. Anthropologically, this is the shape of a living, everyday religiosity rather than a centralised, clerical one: authority radiates outward from thousands of small sacred points, not down from a single institution. It is precisely this decentralised, saint-centred, syncretic texture that made Sindhi Islam so absorptive — able, as Ansari and Schimmel both stressed, to take in the river-god Jhulelal, to share courtyards with Hindus, and to bind ruler to ruled through the mediating figure of the pir.

 

Coda: The Landscape as an Argument

Let me gather the threads, because we have travelled a long way from that first dusk on the river. What has a map of 624 dots actually taught us?

It taught us, first, that geography is destiny: sanctity in Sindh follows water, because life follows water, and where the river moved, the sacred moved with it.

It taught us, second, that the sacred landscape has a birthday — or rather a long, steep adolescence beginning around 1700, when indigenous Muslim dynasties, a maturing class of hereditary saints, and then colonial urbanisation together multiplied holy places faster than any period before.

It taught us, third, that a religious map is a power map. Buddhist ruins mark an ancient conversion; a wall of active Muslim shrines marks the triumph of the pirs; a scatter of colonial churches marks the imperial city; a suspicious absence of surviving urban temples marks 1947. Sites appear when a community rises and fall silent when it is displaced.

And it taught us, finally, that plurality is the deep grammar of Sindh, stubborn even under pressure. The same few kilometres of Karachi still hold a shrine, a temple, a church, a gurdwara, and a fire-temple. Odero Lal still splits its sanctity down the middle. The database, for all its gaps, keeps insisting on this: Sindh has always been a place where many gods shared one floodplain.

Read rightly, a map of graves is a map of the living: of who came, who converted, who prospered, who was driven out, and who — against every pressure — stayed. That is what these 624 dots are really recording.

So the next time someone tells you Sindh is simply ‘the land of Sufis,’ show them this map. It is the land of Sufis — but it is also the land of stranded stupas and desert temples, of colonial churches and merchant gurdwaras. It is a palimpsest written and overwritten for two and a half thousand years, in domes and cupolas, along the banks of a wandering river.

 

On the Data & Its Limits

This article analyses a refined database of 624 geo-located religious sites across Sindh’s 30 districts, each carrying (where verifiable) a religion, architectural type, century of foundation, size, and functional status. It is emphatically not a census. It is an archive of the recoverable: sites whose location, date, and identity could be independently confirmed. Hundreds of further sites were excluded for want of verifiable data. Two biases follow directly and have been flagged throughout: a recency bias (recent, surviving buildings are over-represented against vanished medieval ones) and a survivorship bias (active sites are easier to document than demolished or converted ones, which especially undercounts post-1947 losses of minority sites). Charts of ‘foundations by century’ therefore describe the documented landscape, not the total historical one. Within those limits, the dataset is large and well-distributed enough to reveal the province’s deep sacred structure with confidence. District labels follow the database’s own administrative scheme, including the several sub-districts of Karachi.

Works & Sources Consulted

Ansari, Sarah F. D. (1992). Sufi Saints and State Power: The Pirs of Sind, 1843–1947., Cambridge University Press.

Ansari, Sarah F. D. (2005). Life after Partition: Migration, Community and Strife in Sindh, 1947–1962., Oxford University Press.

Kalhoro, Zulfiqar Ali (2022). Saints, Sufis and Shrines: The Mystical Landscape of Sindh., Endowment Fund Trust for Preservation of the Heritage of Sindh

Schimmel, Annemarie (1986). Pearls from the Indus: Studies in Sindhi Culture., Sindhi Adabi Board, Jamshoro.

Boivin, Michel (2007), Sindh through History and Representations: French Contributions to Sindhi Studies. Oxford University Press.

Boivin, Michel (2019), The Hindu Sufis of South Asia: Partition, Shrine Culture and the Sindhis in India., IB Tauris and Co Ltd.

Boivin, Michel, and  Pénicaud, Manoël (2023) Inter-religious Practices and Saint Veneration in the Muslim World: Khidr/Khizr from the Middle East to South Asia., Routledge

Dalrymple, Sam (2026). Shattered Lands; and essays ‘Cosmopolitan Karachi’ and ‘Relics of Partition’, Travels of Samwise.

Migration Affairs Journal — ‘Some Fruits of Freedom: Partition and the History of Evacuee Property in Pakistan.’

Singh, Bhajan (2019). Hindu Temples in Pakistan: During Partition and Aftermath., (survey of Evacuee Trust Property Board sites).

Lajwani, Dr. Ali Murad (2019), A Unique Study of Sufi Rituals at the Shrines of Sufi Saints of Pakistan, ethnographic study of shrine practice in Sindh. Al Khadim Research Journal of Islamic Culture and Civilization

Hasan, S. Khurshid; Hastenrath, S. Zajadacz — studies of Chaukhandi tombs and Makli funerary architecture, lower Sindh.

Primary contextual sources: the  Chach Nama and the travelogue of Xuanzang (Hiuen Tsiang) on pre-Muslim Sindh.

Endowment Fund Trust for the Preservation of the Heritage of Sindh (EFT) & Government of Sindh heritage listings — site-level documentation underlying many database entries.