Why the district was expected to go to Pakistan

This wasn't a rumour — it was the operative legal position on 14 August. Under the 3 June Plan and the Indian Independence Act, a notional boundary governed the transfer of power until the Boundary Commission's award superseded it. That notional line assigned Muslim-majority districts of the Lahore and Rawalpindi divisions to West Punjab. Gurdaspur, in Lahore Division, had a bare Muslim majority in the 1941 census (a little over 50% district-wide), and three of its four tehsils were Muslim-majority: Batala at about 55%, Gurdaspur at 52%, Shakargarh at 51%, with only Pathankot clearly non-Muslim. So the district fell to Pakistan notionally, and the administrative machinery followed. The News

Hence the detail that anchors the whole episode: Pakistan sent Mushtaq Ahmed Cheema as Deputy Commissioner of Gurdaspur, and the Pakistan flag flew over the district for those days. The incumbency board in the DC's office still lists Cheema as DC from 14 to 17 August 1947; Chanaya Lal took over from him on the 17th. That board is the single most-cited physical trace of the interregnum. The Partition Museumtribuneindia

14–16 August: the vacuum

Radcliffe finished the Punjab and Bengal awards around 9–12 August and they were with Mountbatten by the 13th. He held them. The stated reason was to avoid confusion on independence day itself; the less flattering reading is that Mountbatten did not want the boundary question overshadowing the ceremonies in Karachi and Delhi. He released the awards to Indian and Pakistani leaders on the afternoon of 16 August, and they went out publicly on the 17th — which is why you'll see both dates in the literature. The 16th is the leaders' meeting; the 17th is the broadcast and publication. The DiplomatPunjab Partition

For Hindus and Sikhs in Gurdaspur, Batala, Qadian and the Ravi belt, those three days were lived as confirmation that they were now in Pakistan. Sikh landholding in Batala tehsil was substantial, Qadian was the Ahmadiyya centre, and the district also held one of Punjab's largest Christian populations. The reaction was the movement you'd expect: flight eastward toward Pathankot and across the Beas, and toward Amritsar — which was itself under attack. Sikh jathas were already mobilised across the Amritsar–Gurdaspur belt through the second week of August, and the perception that the district was being handed over hardened that. Master Tara Singh and the Akali leadership had been agitating for Gurdaspur specifically; the Sikhs were eyeing Gurdaspur to be included in India, and its apparent loss read as the final betrayal after Lahore, Lyallpur and Nankana Sahib. University of the Punjab

17 August: the reversal

The award gave the three tehsils east of the Ravi — Gurdaspur, Batala and Pathankot — to India, while Shakargarh, west of the river, went to Pakistan. Cheema left for Pakistan the same day; the tricolour replaced the Pakistani flag. The Partition Museum

The relief on one side was immediate and the panic simply inverted. Muslims in Batala, Gurdaspur and Qadian — a majority in two of those tehsils — went overnight from being the notional majority community of a Pakistani district to a minority in India, with no time to have moved and with the retaliatory logic of the Amritsar violence now extending north. Batala was a substantial Muslim commercial town; its emptying was one of the sharper reversals in East Punjab. The mass columns moved through late August and September, and the Ravi crossings became the killing grounds: the Dera Baba Nanak crossing over the Ravi, like Balloki head near Lahore, became a place where mobs waited for caravans to halt at the bridge. Punjab Partition

The Punjab Boundary Force was nominally responsible for the district and was manifestly unable to hold it; it was wound up on 1 September.

Reception and the argument that followed

Jinnah's line on the award was that Pakistan had been squeezed as hard as was possible, and Gurdaspur became the central Pakistani grievance alongside Ferozepur and Zira. The objection was straightforward: Gurdaspur and Batala tehsils had Muslim majorities and were contiguous with West Punjab, yet went east. The Kashmir argument followed almost immediately — that Pathankot gave India its only land route to Jammu, which mattered enormously by October 1947. Punjab PartitionPunjab Partition

Radcliffe's own justification rested on the Upper Bari Doab Canal: the Madhopur headworks in Pathankot fed the canal system irrigating Amritsar district, and severing an admittedly non-Muslim Amritsar from its water source was the anomaly he chose to avoid. Whether that was reasoning or rationalisation is the argument that has never closed, and the Abell note of 8 August — showing an earlier line more favourable to Pakistan — keeps it open.

Most of the granular violence in the district is dated from roughly 18 August onward. The three days are better characterised as an administrative interregnum that set the conditions, rather than as a discrete massacre window.

The district split into two violence regimes

The Ravi is the analytical key. Shakargarh tehsil, west of the river, went to Pakistan; Gurdaspur, Batala and Pathankot, east of it, went to India. From roughly 17–18 August the killing in the two halves ran in opposite directions, and the river crossings between them became the worst ground of all.

Shakargarh tehsil — attacks on Hindus and Sikhs

Shakargarh had a slim Muslim majority (~51%) with substantial Sikh landholding, and it lay against Sialkot and Narowal. The Punjab government's compiled massacre lists (the material behind Khosla's Stern Reckoning and Talib's SGPC compilation) place a cluster of attacks here in the second half of August:

• Jhorata, Dhamray and Bhatie — from 17 August onward, Sikhs killed by Muslim mobs acting with Baloch troops.

• Bhawanipura — 24 August, Hindus and Sikhs attacked, survivors hiding in standing sugarcane.

• Viran Datta and Kotla Pathanan — 24 to 26 August, villages attacked and looted, killings, large numbers of women abducted, forced conversions.

• Gurdwara Bali Sahib — attacked by a mob of several hundred, the building burned with people inside.

The pattern is the standard West Punjab one after the award: police and demobilised soldiery participating or standing aside, attacks announced by drums, villages cleared rather than merely raided.

The Ravi crossings — Dera Baba Nanak, Jassar, Ghanike Bet

Because Shakargarh, Narowal and Sialkot emptied eastward and the Batala–Gurdaspur belt emptied westward, both flows funnelled to the same handful of bridges and fords. These became deliberate ambush points: mobs waited at the crossings for caravans to halt and attacked them there, with bodies going into the river — the same tactic used at Balloki head near Lahore.

Kirpal Singh's testimony is one of the more usable first-person accounts for your purposes: his village was Chandiawala in Narowal, his family left on 20 August, and at Ghanike Bet inside Gurdaspur he witnessed the full repertoire — women drowning themselves in the Ravi, abductions, and killings of women by their own families.

The Jassar massacre of 23 October 1947 belongs here too. Jassar was the Sialkot-side railway junction on the Ravi; a column of non-Muslim refugees moving on foot toward Dera Baba Nanak was attacked there. Pakistani press reported roughly 30 dead; survivor accounts circulating in Amritsar put it at 1,500. The Deputy Commissioner of Gurdaspur went to Dera Baba Nanak with the brigade commander and SP to investigate. The consequence was immediate: about 10,000 non-Muslims massed at the Cattle Fair ground camp in Amritsar intending to attack the Muslim refugee camp there. That's a clean documented example of how a killing on the Pakistani bank produced retaliation pressure on the Indian bank within a day.

Batala, Gurdaspur and Dhariwal — attacks on Muslims

Batala tehsil was 55% Muslim and Batala town was a significant Muslim commercial centre. When the award reversed the district's status on 17 August, that population became a stranded minority in a district already full of armed Sikh jathas and incoming refugees from Sialkot and Narowal carrying accounts of what had been done to them.

The reprisals in the eastern tehsils were sustained rather than episodic — carried out over days by combinations of villages numbering in the thousands, on a scale where burial of the dead became a logistical problem. Batala town was one of the worst sites; accounts of people detained there in transit describe conditions in the town during late August and September as sustained horror. Dhariwal (the mill town), Dinanagar, Kalanaur and Fatehgarh Churian all lost their Muslim populations in this period.

Gurdaspur's large Christian population — unusually large by Punjab standards, concentrated around Batala's mission institutions — largely stayed put, and mission compounds functioned as refuges in places.

Qadian — the best-documented single site

Qadian, in Batala tehsil, is the one location in the district with a near-daily contemporaneous record, because Mirza Bashir Ahmad kept a diary of the siege. It is worth treating as its own entry.

Qadian was the Ahmadiyya headquarters, Muslim-majority, and now marooned about fifteen miles inside India. Ahmadis from surrounding villages converged on the town — figures given in Ahmadiyya sources run to tens of thousands, and should be treated as estimates from an interested party. What the diary and corroborating material record:

• Bodies in the streets of the town and its surroundings.

• Makeshift refugee camps across the town.

• Looting and ransacking of Muslim property, including the house of Sir Muhammad Zafrullah Khan.

• Bomb and grenade attacks on Masjid Aqsa, the central mosque, including a grenade thrown into the mosque precincts on 10 October 1947.

• The departure of the Khalifa, Mirza Mahmud Ahmad, for Lahore at the end of August, after which a small group drawn by lot stayed behind to hold the sacred sites.

  
  

The external verification matters for citation purposes: a British representative named Stephenson visited Qadian on 11 October 1947 and confirmed the town was under continued heavy attack involving murder, looting and destruction; the Civil & Military Gazette covered it on 23 October; and two British journalists, Douglas Brown and Josselyn Hennessy, went to Qadian in October specifically to investigate. Zafrullah Khan also approached Gandhi to get an observer sent. Death tolls for Qadian in Ahmadiyya records are surprisingly modest — a few dozen — which tells you the defence largely held even as the town was destroyed and emptied. Claims circulating about armoured vehicles or organised Ahmadi militia at Qadian come from polemical sources on both sides and I'd treat them as unverified.

Institutional context

The Punjab Boundary Force covered Gurdaspur among its twelve districts and was demonstrably unable to hold it; it was dissolved on 1 September, leaving the district to the new East Punjab administration and the Military Evacuation Organisation. Evacuation of Muslims from Gurdaspur and Batala ran through September and October, mostly by foot column to Dera Baba Nanak and by rail via Amritsar — both routes with documented attacks.

Shakargarh tehsil — 17 to 26 August

Jhorata, Dhamray and Bhatie, from 17 August onward. Sikhs killed by Muslim mobs acting together with Baloch military. The compilers add an allegation worth flagging separately in any entry: that British officers of the Punjab Boundary Force were assisting the attackers. That is a partisan claim from an interested source, but it recurs and it bears on why the PBF collapsed in credibility.

Bhawanipura, 24 August. Hindus and Sikhs hid in standing sugarcane. Assurances of safety were given to draw them out; those who emerged were killed. A handful of women and three men survived initially, and the men were killed later. Young girls abducted.

Viran Datta and Kotla Pathanan, 24 to 26 August. Mob and military together. Killings, looting, large-scale abduction of women, forced conversions.

Bhopalwala (just over the Sialkot line, same belt), end of August: Gurdwara Bali Sahib attacked by around 400 men, the building burned with people inside.

The pattern across all of these is consistent: attack announced by drums, participation by police or regular troops, false assurances used to concentrate victims, conversion offered and then reneged on.

The Ravi crossings

This is where Gurdaspur's geography did the killing, and where the two flows collided.

Narowal, 17 August. Looting and murder in the town; those who fled were massacred while crossing the Ravi bridge, with Sikhs singled out. This is the northern feeder into the Dera Baba Nanak route.

Bhikho Chak, 24 August. A party of about 130 attacked Hindus and Sikhs leaving the village. The attack was made at the pattan on the Ravi. 113 killed, 17 or 18 survivors.

Badomalli, 29 August. An attacking force reported at 6,000 using rifles, machine guns and artillery pieces, with drums. Around 3,000 Hindus and Sikhs killed, 300 women abducted, survivors forcibly converted. This is the largest single figure in the Ravi belt and should be treated as a claimed figure, not a verified one.

Ghanike Bet, around 20 August. Kirpal Singh, whose village was Chandiawala in Narowal, crossed into Gurdaspur here and described women throwing themselves into the Ravi, abductions, and women killed by their own families.

Jassar, 23 October. A foot column of non-Muslims moving toward the Dera Baba Nanak crossing was attacked at the Sialkot-side railway junction. Pakistani accounts gave roughly 30 killed and 200 wounded; survivor accounts in Amritsar put it at 1,500. The DC of Gurdaspur went to Dera Baba Nanak with the brigade commander and SP. Within a day about 10,000 non-Muslims assembled at the Cattle Fair ground in Amritsar intending to attack the Muslim refugee camp there. That sequence — killing on the west bank, retaliation threatened on the east bank within 24 hours — is the district's characteristic mechanism.

East of the Ravi — the reprisal, from roughly 18 August

The most direct account of what happened in the three Indian tehsils is Kirpal Singh's, describing the arrival of refugee Sikhs at Thetherke in Gurdaspur. What he describes is not a riot but a coordinated campaign: several thousand men from a wide radius of villages grouping together, sustained over days, on a scale where burial capacity ran out and the wounded were left where they fell because no one would approach. Jatha leaders announced their instructions before attacks — kill the males, spare the women. That instruction is not mercy; it identifies abduction as policy.

Batala town, Batala tehsil generally, Dhariwal, Dinanagar, Kalanaur and Fatehgarh Churian all lost their Muslim populations in this phase. Batala is the site with the most independent corroboration, because outsiders passed through it: one Ahmadi traveller detained there for two days in this period recorded what he witnessed, and the town appears repeatedly in accounts as the worst place in the district.

There is a structural reason East Punjab's version is thinner in the record. The Punjab government compilations document only what was done to Hindus and Sikhs; the Pakistani side had no equivalent access to East Punjab villages to compile. So Batala and Gurdaspur tehsils, where the killing after 17 August was probably heaviest in the district, have the least village-level documentation. That asymmetry is itself worth stating in your atlas rather than letting it silently shape the map.

Qadian, September–October

Qadian is the exception — the one Indian-side location with a near-daily contemporaneous record, kept by Mirza Bashir Ahmad. The town held out far longer than the rest of the district's Muslim settlements.

• Late August: refugees converge from surrounding villages; Ahmadiyya sources give figures in the tens of thousands.

• Bodies in the streets and outskirts; makeshift camps across the town.

• Looting and ransacking of Muslim property, including Sir Muhammad Zafrullah Khan's house.

• A student on guard duty at Talim-ul-Islam College killed.

• End of August: the Khalifa leaves for Lahore. A small group, selected by lot, stays to hold the sacred sites.

• 14 September: arrest of a senior Ahmadiyya office-holder by the Indian authorities — the administration was detaining the besieged as well.

• A man who went out for a morning walk on a Friday disappeared and was presumed killed.

• Delegations from the community were made to wait for a DSP coming from Batala and then not received.

• 10 October: a hand grenade thrown into the Masjid Aqsa precincts, following an earlier bomb.

• 11 October: a British representative, Stephenson, visits and confirms continued heavy attack with murder, looting and destruction. The Civil & Military Gazette reports it on 23 October. Douglas Brown and Josselyn Hennessy visit as journalists in October. Zafrullah Khan had approached Gandhi to get an observer sent.

  
  

Recorded Ahmadi deaths at Qadian are modest — a few dozen — which is the useful datum: the defence held even as the town was emptied. Claims about armoured vehicles or a large Ahmadi militia come from polemical sources and shouldn't be plotted as fact.

Two cautions for plotting

First, the large round figures (3,000 at Badomalli, 1,500 at Jassar) come from compilations gathered from survivors under conditions that produced systematic upward drift. Plot them as claimed with the source attached.

Second, several sites in these lists are attributed to "Muslim mob and Military" or "Muslim mob, Police and Military." Whether that means regular Pakistan Army units, the Baloch battalions of the disintegrating PBF, or police is often not resolvable from the entry. The distinction matters and is usually not recoverable, so it's better recorded as ambiguous than resolved by inference.

The short answer is that Muslims in Gurdaspur were not killed by riots. They were killed by an organised clearance campaign, run by armed formations, in four recurring situations. Here's how it worked.

The instrument: the jatha

The jathas operating in the Gurdaspur–Amritsar belt were paramilitary bands, not crowds. They carried Bren guns, Tommy guns, grenades, rifles, revolvers, pistols, spears and kirpans. Many were motorised, moving between targets on trucks and jeeps. Their ranks included Sikhs displaced from Rawalpindi Division in March, who had come east specifically to take reprisal, and demobilised soldiers with wartime training.

Jenkins was reporting jatha attacks on Muslim villages around Amritsar to the Viceroy's House as early as 1 August — roughly two dozen Muslims killed and thirty wounded inside 48 hours, and four passenger trains attacked — with the assessment that a larger offensive was being prepared. That offensive was launched after 15 August, and Gurdaspur entered it on 17–18 August, when the award removed the last reason to hold back.

The Times correspondent, writing on 25 August, described the clearance of eastern Punjab as systematic and worked through sector by sector, organised at the level of Sikh leadership rather than locally improvised. He flew over Jullundur district and counted around fifty villages burning in a two-hour reconnaissance. British and Indian officers told him it was worse than anything they had seen in the war. Amritsar and Jullundur towns had gone quiet by then for the simple reason that no Muslims remained in them.

The four killing situations

1. Village clearance. The standard method: surround a Muslim village, usually at dawn, with drums beaten during the advance. Men and boys killed. Houses fired. Women and girls taken. Kirpal Singh's account of Gurdaspur is explicit about the instruction being announced to the jatha before the attack — kill all the males, spare the women. That is not restraint; it identifies abduction as the intended outcome for women rather than an incidental one. The same instruction structure appears across East Punjab accounts.

In Gurdaspur specifically, Kirpal Singh describes what happened after refugee Sikhs from Narowal reached Thetherke: the impulse to revenge spread across the whole district, several thousand men from different villages combining, and attacks sustained over days rather than hours. His detail about burial capacity being exhausted, and the wounded left calling for help with nobody willing to approach, is the clearest indication of scale we have for the Indian tehsils.

2. The concentration, then the attack on it. Muslims were collected — in a mohalla, a mosque, a school, or a designated camp — often on an assurance of safety from officials or the escort, and then attacked once concentrated. This is the same technique used against Sikhs at Sheikhupura and Jaranwala in the same weeks, running the other direction. It produced the highest death counts on both sides because it removed the possibility of dispersal.

3. The foot column. This was probably the largest killer for Gurdaspur's Muslims. Columns moving west toward the Ravi were slow, strung out over half a kilometre or more, encumbered with carts, and escorted by police or troops whose communal loyalty was frequently the problem rather than the protection. The documented pattern in East Punjab convoy massacres is that the escort withdraws or steps aside at a prearranged moment and armed parties emerge from the adjacent fields. Columns were also refused food and water in villages they passed through, which forced halts, and halts were where attacks came.

The killing concentrated at bottlenecks: river crossings and ferry points, bridges, canal bunds, railway level crossings, and the approaches to towns. For Gurdaspur that meant the Ravi crossings — the same points where columns moving the other way were being killed by Muslim mobs. Dera Baba Nanak was a kill point for both directions.

4. The train. Trains carrying Muslims west from Gurdaspur and Batala ran through Amritsar, which was the worst single stretch of track in Punjab. In September, jathas killed around a thousand Muslim refugees on a Pakistan-bound train near Khalsa College, Amritsar. On 22 September a train leaving the Amritsar area was attacked with roughly 3,000 killed. Mudie recorded that in the sequence of attacks on westbound trains from Amritsar and Jullundur districts between 21 and 23 September, one train had every occupant killed. Pakistan suspended all trains from Indian Punjab at the end of September as a direct result.

The Amritsar bottleneck

Gurdaspur's Muslim columns had to get past or through Amritsar, and Amritsar was effectively closed to them. Jathas of up to 10,000 attacked convoy escorts; in one instance such a force attacked a detachment of sixty soldiers and wounded their Sikh officer four times. Commanders had a further problem: Sikh troops were expected to desert rather than keep firing on their own community. At a conference with Liaquat Ali Khan in Lahore, Nehru had to concede that his government could not open Amritsar to Muslim convoys. Thimayya's fallback was to bulldoze a bypass road around the city, which took days. The refugees waited in the open in the meantime, attacked repeatedly and denied food and water.

Abduction

Abduction was systematic, not opportunistic. The instruction to spare women meant taking them. The recovery operation set up afterwards under the Inter-Dominion agreement was still recovering abducted women from both Punjabs into the 1950s, and Gurdaspur and Batala figure in those recovery caseloads. Women killed by their own male relatives to prevent abduction is a recurring feature on the Muslim side of Gurdaspur just as it is on the Sikh side in Shakargarh — the same act, in mirror image, in the same district within a fortnight.

Numbers, and why they don't exist for this district

There is no Gurdaspur equivalent of the Sheikhupura enquiry for the Muslim dead. The reason is structural: the Punjab government compilations that give village-level detail were compiled to document what was done to Hindus and Sikhs in West Punjab. Pakistan had no comparable access to East Punjab villages after August, and the Pakistani documentation of East Punjab is thinner and more heavily contested. Overall estimates for Muslim deaths in eastern Punjab as a whole run into the hundreds of thousands, but no defensible district figure for Gurdaspur exists.

A note on sources

Nearly all the village-level detail for this district comes from partisan compilations: the Punjab government and SGPC lists on one side, Pakistani official reports and Ahmadiyya records on the other, each documenting its own community's losses and silent on its own community's actions. For an atlas, the defensible move is to plot the sites and attribute the claim to its source rather than adjudicating counts. Gurdaspur has no equivalent of the Sheikhupura or Rawalpindi enquiries — no single authoritative reckoning — and district-level death estimates for it are not reliable.