The Pakistan Army’s Inter-Services Public Relations has scorned Indian army's claims that it conducted a “surgical strike” across the Line of Control that divides the erstwhile princely state of Jammu and Kashmir between India and Pakistan on the night of September 28-29.

“No such incident took place nor will we allow any such incident to happen in future,” insisted Lt Gen Asim Saleem Bajwa.

Pakistan’s case appears to be built on two things.

  1. Neither widely quoted (as per reports in the Washington Post or BBC) residents from local areas nor the United Nations Military Observer Group in India and Pakistan observers saw anything.

  2. It simply is not possible for India to cross the LoC. Bajwa claims (video in Urdu) “there have never been physical violations, nor can it happen, nor would we allow it”.

The first point was flawed, given the Army’s influence over villages near the LoC (see Ayesha Siddiqa here), the acknowledged large-scale use of diversionary shellfire, and the geography of LoC – nevertheless, it has been taken on thoroughly by Praveen Swami (whose reporting also provides the bulk of material below) in the Indian Express of October 5 through eyewitness reports.

But the point of this post is to demonstrate that Bajwa is flat-out wrong on the second.

What follows is a cursory and no doubt very partial history of allegations of raids across the LoC. The point is that they are not unprecedented, and that the LoC is not sacrosanct at a reasonably low level of conflict.

I can find one reference to a pre-1998 raid in an almanac that notes that on September 1, 1991:

″Indian troops attack a Pakistani post in Nezarpur on the Line of control, killing three Pakistani soldiers” (“Pakistan 1992“, Westview Press, December 1992, p180). 

I can locate Nazar Pur (presumably the same) in Bagh district of Pakistan-occupied Kashmir, though it looks to me to be about 10 kms across the nearest point on the LoC and therefore an unlikely candidate. I can find no other reference to this incident, though perhaps a more thorough search of the Pakistani newspaper archives would throw something up.

But several years later, in 1998, the Pakistan Army itself acknowledged an Indian raid, in a complaint it issued to the United Nations in 1998 (Yearbook of the United Nations 1998, p321):

Pakistan reported on 4 May the massacre of 22 Kashmiri civilians in the village of Bandala in Azad Kashmir, approximately 600 metres from the line of control on the side of Azad [i.e. Pakistan-occupied] Jammu and Kashmir. An investigation had established beyond reasonable doubt, said Pakistan, that the massacre … "was an act of terrorism by Indian armed forces who had crossed the line of control”.

The Bandala incident was widely reported at the time. This is Dexter Filkins in the Los Angeles Times in May 1998:

Last month, in a widely reported incident, 22 villagers were massacred in Bandala Seri by gunmen believed to be from India. The villagers say the gang of a dozen men, all dressed in black, struck in the middle of the night and dropped leaflets to mark the attack. “Vengeance Brigade,” one leaflet said. “Evil deeds bear evil fruit,” said another. “Ten eyes for one eye, one jaw for a single tooth,” said a third. Blood still stains the walls of several huts. When the Pakistani government accused the Indian government of sponsoring the attack, New Delhi denied any responsibility. Some, including U.S. officials, believe the attack may have come in retaliation for the killing of 26 Indian civilians a week before in the villages of Parankot and Dhakikot.

Praveen Swami wrote about it in July of the same year, in a fascinating piece for Frontline magazine:

“Frankly,” says one Army official, “I’m amazed.” “We could, for example, take out the five major launching posts for terrorists in the Lipa-Jura arc just across the LoC, but it wouldn’t take Pakistan very long to set up a few tents and get going again. We could raid the Muzaffarabad training camps but there would be large-scale casualties, including civilian deaths, and that would be diplomatically unacceptable.” Other options, including reprisals against Pakistan, also appear ill thought-out. The April 27 massacre of 21 villagers in Binda Mohri Sehri, metres across the LoC inside Pakistan-Occupied Kashmir, and the bombing in June of a Lahore-bound train, shortly after the explosion in Jammu, are both believed by Pakistan to have been carried out by Indian security agencies. “If your intelligence apparatus,” argues one official, “encourages anti-Pakistan forces in Gilgit and Hunza, which it is not doing, that has a strategic purpose. Mindless retaliatory killing simply fuels domestic legitimacy for what Pakistan is doing over here, and gives no returns.”

Then there’s a 15-year gap, but Swami – ever diligent – returns in The Hindu with even clearer sources, and a wealth of information on more Indian raids.

The most savage cross-LoC violence Indian forces are alleged to have participated in was the killing of 22 civilians at the village of Bandala, in the Chhamb sector, on the night of March 26-27, 1998. The bodies of two civilians, according to Pakistan’s complaint to UNMOGIP, were decapitated; the eyes of several others were allegedly gouged out by the attackers. The Pakistani military claimed to have recovered an Indian-made watch from the scene of the carnage, along with a hand-written note which asked, “How does your own blood feel”? First reported by The Hindu’s sister publication Frontline in its June 19, 1998 issue, the Bandala massacre is alleged to have been carried out by irregulars backed by Indian special forces in retaliation for the massacre of 29 Hindu villagers at Prankote, in Jammu and Kashmir, by the Lashkar-e-Taiba. The LeT attackers slit the throats of their victims, who included women and children. No Indian investigation of the Bandala killings has ever been carried out. However, an officer serving in the Northern Command at the time said the massacre was “intended to signal that communal massacres by jihadists, who were after all trained and equipped by Pakistan’s military, were a red line that could not be crossed with impunity.”

Then a couple of weeks ago, Rahul Singh and Rajesh Ahuja in the Hindustan Times add an intelligence source to Swami’s military one:

Kashmir watchers recall one such raid that took place after 26 Hindus were killed in Parankote and Dhakikote villages in Udhampur in April 1998. “Intelligence inputs suggested that the massacre was carried out by the Lashkar terrorists. We retaliated,” said a former IB official on condition of anonymity. He, however, did not specify the exact nature of retaliation. But a few weeks later, Pakistan claimed 22 people were killed when unknown gunmen attacked its Bandala Seri village.

It’s fair to say that when Indian military and intelligence personnel are owning up to gruesome atrocities, to reliable journalists in good newspapers, we can take this seriously.

Then there’s a host of other examples.

Ajai Shukla writing in 2013, on an example from the summer of 1999 during the Kargil conflict:

But the [Loc] fence must be physically monitored and so, small groups of Indian soldiers patrol the gaps between posts. These “area domination patrols” are particularly vulnerable whilst in the sliver of Indian territory ahead of the LoC fence. In Poonch, the patrol was ambushed ahead of the fence. For carrying out such ambushes and for attacking small enemy posts in tactically favourable terrain, both India and Pakistan have contingency plans worked out and rehearsed. When one side needs to send a signal, or to retaliate, one of those plans is implemented at short notice. During the Kargil conflict in the summer of 1999, an Indian Army platoon had crossed the LoC at the Munawar Tawi River near Jammu and wiped out an entire Pakistani post, triggering a vicious cycle of revenge killings and counter-killings. At that time, Pakistan refined the concept of “border action teams”, or BATs, specifically earmarked for sneak killings along the LoC. Some BATs feature commandoes from Pakistan’s elite Special Services Group (SSG), while others are constituted from local forces. The beheading of an Indian soldier in January has been ascribed to a BAT.

Returning to Swami’s 2013 Hindu piece:

Six months after the Kargil war, on the night of January 21-22, 2000, seven Pakistani soldiers were alleged to have been captured in a raid on a post in the Nadala enclave, across the Neelam River. The seven soldiers, wounded in fire, were allegedly tied up and dragged across a ravine running across the LoC. The bodies were returned, according to Pakistan’s complaint, bearing signs of brutal torture. “Pakistan chose to underplay the Nadala incident,” a senior Pakistani military officer involved with its Military Operations Directorate told The Hindu, “as General Pervez Musharraf had only recently staged his coup, and did not want a public outcry that would spark a crisis with India.” Indian military sources told The Hindu that the raid, conducted by a special forces unit, was intended to avenge the killing of Captain Saurabh Kalia, and five soldiers – sepoys Bhanwar Lal Bagaria, Arjun Ram, Bhika Ram, Moola Ram and Naresh Singh – of the 4 Jat Regiment.

There is general agreement that such raids slowed after the 2003 ceasefire was agreed between India and Pakistan (“pre-2003 this was routine stuff”, says Indian Express journalist and Indian Army veteran Sushant Singh), but Swami finds further examples:

Less detail is available on the retaliatory cycles involved in incidents that have taken place since the ceasefire went into place along the LoC in 2003 – but Pakistan’s complaints to UNMOGIP suggest that there has been steady, but largely unreported, cross-border violence involving beheadings and mutilations. Indian troops, Pakistan alleged, killed a JCO, or junior commissioned officer, and three soldiers in a raid on a post in the Baroh sector, near Bhimber Gali in Poonch, on September 18, 2003. The raiders, it told UNMOGIP, decapitated one soldier and carried his head off as a trophy. Near-identical incidents have taken place on at least two occasions since 2008, when hostilities on the LoC began to escalate again. Indian troops, Pakistan’s complaints record, beheaded a soldier and carried his head across on June 19, 2008, in the Bhattal sector in Poonch. Four Pakistani soldiers, UNMOGIP was told, died in the raid. The killings came soon after a June 5, 2008, attack on the Kranti border observation post near Salhotri village in Poonch, which claimed the life of 2-8 Gurkha Regiment soldier Jawashwar Chhame. Finally, on August 30, 2011, Pakistan complained that three soldiers, including a JCO, were beheaded in an Indian raid on a post in the Sharda sector, across the Neelam river valley in Kel. The Hindu had first reported the incident based on testimony from Indian military sources, who said two Pakistani soldiers had been beheaded following the decapitation of two Indian soldiers near Karnah. The raid on the Indian forward position, a highly placed military source said, was carried out by Pakistani special forces, who used rafts to penetrate India’s defences along the LoC.

In a separate piece in The Hindu, Swami furnishes another example from January 2013:

Finally, on January 6 [2013], matters came to a head. Following a low-grade exchange of fire that night, 19 Infantry Division commander Gulab Singh Rawat sought and obtained permission for aggressive action against the Pakistani position from where his troops were being targeted. Pakistan insists its post, Sawan Patra, was raided by Indian troops. India has denied the allegation. “None of our troops crossed the Line of Control,” said Jagdish Dahiya, an Indian army spokesperson. Either way, though, a Pakistani soldier was dead before the shooting ended – and another critically injured.“Let’s just put it this way,” a senior government official in New Delhi said, “there was no formal permission to stage a cross-border raid to target Sawan Patra. However, in the heat of fighting, these things have been known to happen. Pakistan has done this, and our forces have done this, ever since fighting began in Jammu and Kashmir in 1990.”

And yet another example by Swami in First Post, this time from August 2013:

[E]arly this month, Zafran Ghulam Sarwar, Wajid Akbar, Mohammad Wajid Akbar and Mohammad Faisal left their homes on the Pakistani side of the control in the Neelam valley, and never came back … India says it has no idea what happened to the men. Not long after they disappeared, though, five still-unidentified men were shot dead by Indian troops in the same area, 500 metres on the Indian side of the Line of Control. Naresh Vij, an Indian army spokesperson, said troops had “not recovered any bodies as they are lying very far.”Privately, Indian intelligence officials posted in the sector speculate the men may have indeed been targetted by special forces – but insist they were guides for jihadist groups crossing the Line of Control, not innocent men executed by the army for no reason at all.

Saikat Datta, writing a tick-tock of the latest raid for Scroll, makes reference to the older raids to draw a contrast with the latest:

These surgical strikes are not the first of their kind that India has carried out across the LoC. While the Vajpayee government took a clear decision not to allow the Indian forces to cross the LoC during the Kargil war, Indian Special Forces carried out several raids between 2000 and 2003. However, after the November 2003 ceasefire, such raids had been called off. Some Special Forces raids were renewed after 2012 when Pakistani troops raided Indian army posts and beheaded Indian soldiers in some cross-LoC operations. Most of the earlier operations were local military actions, held either at the Corps or Divisional level, with local intelligence units being assigned with gathering information before strikes. This time, the R&AW was specifically tasked by the National Security Advisor to gather actionable intelligence, which could be used to carry out pinpoint strikes with minimum casualties. A conscious decision was also taken to target militants rather that the Pakistani army to ensure that the situation did not escalate …“We have had strikes earlier, but those were mostly local,” Lieutenant General Hardev Singh Lidder, a former Chief of Integrated Defence Staff and a veteran Special Forces officer told “This is the first time that strikes were carried out as a national policy, which is significant.”

In a separate piece also for Scroll, Datta goes as far as to give us a specific example authorised by then Prime Minister Atal Behari Vajpayee in 2000, conducted by India’s 9 Para (Special Forces):

Between 1998 and 2014, there have been several strikes by Indian forces across the LoC. Just after the Kargil war, Captain Gurjinder Singh Suri, posted on the LoC with 12 Bihar battalion took a team of ghataks (infantry battalion commandos) across the LoC to take out Pakistani posts in retaliation of an earlier attack. While Captain Suri was killed in the assault, he was posthumously awarded the Maha Vir Chakra, India’s second-highest military gallantry award. On March 2, 2000, Lashkar-e-Taiba militants massacred 35 Sikhs leading to a major covert operation. A team from 9 Para (Special Forces) was sanctioned by Prime Minister Atal Behari Vajpayee’s government to carry out a raid inside Pakistan. Led by a Major, the Special Forces team went into Pakistan and came back after killing over 28 Pakistani soldiers and militants. The proof of their action was never disputed. Similar raids took place in 2007 and 2013, in retaliation for attacks against Indian military targets.

And then, predictably, on October 4, 2016 the Congress Party decided that it did not want to miss out on the “surgical strikes” party. It released three dates on which it claimed its own cross-LoC assaults, which the Indian Express’ Shubhajit Roy cross-references with public reports of the time:

The Congress on Monday claimed the United Progressive Alliance II, like the National Democratic Alliance government, had also conducted “surgical strikes” – but without making them public. The party listed three dates – September 1, 2011; July 28, 2013; and, January 14, 201 – when the strikes took place. The Manmohan Singh government, known for its publicly stated policy of “strategic restraint”, was in power at the time.

So we have cross-LoC raids alleged in:

  • May 1998
  • Summer 1999
  • January 2000
  • March 2000
  • September 2003
  • June 2008
  • August 2011
  • January 2013
  • August 2013

Only the first of these – May 1998 – is acknowledged by Pakistan officially, but others are acknowledged by Indian officials, albeit anonymously, despite the fact that they represent clear instances of torture (January 2000) or desecration of bodies (June 2008 and August 2011). Unlike last week’s strikes, which were deliberately and carefully framed in terms of pre-emption, these are not admissions that can be written off as Indian attempts to look good. We can safely assume there have been many others, played down by both sides in order to control escalation (not least because the level of political or higher authorisation for some of these appears to have varied).

The LoC has changed in important ways since the late 1990s. On India’s side at least, it has three tiers, remote-controlled machine guns, ground and motion sensors, monitoring by drones, and sophisticated new construction is underway. But these reported strikes appear to have continued well into the last few years, suggesting that improvements on the Pakistani side have not made the LoC impermeable, as Bajwa implausibly claims. Given the geography and terrain, it is unlikely it could ever be made so.

Many of the reported details of India’s September 28-29 strikes are likely to prove inaccurate or exaggerated. But shallow incursions across the LoC by both sides have happened before, and they will happen again. Those dismissing out of hand the September raid on this basis need to pay careful attention to the history.

Shashank Joshi is Senior Fellow at Royal United Services Institute, Associate at Oxford's Changing Character of War Programme and author of Indian Power Projection. A version of this first appeared on his blog.