The lost battles of Khara and Pili

The 10 years of the Maoist conflict in Nepal, 1996-2006, cost just over 13,000 lives. About 8000 were killed by the security forces. Many of those who died were civilians, and some thousands were extra-judicially executed or 'disappeared' – again, the great majority by the security forces. An unquantifiable number of combatants were killed in battles and genuine encounters between the Royal Nepal Army (RNA) and the armed wing of the Maoists, the 'People's Liberation Army' (PLA). About 350 of these individuals were killed in two battles that took place in 2005: at Khara, on 7 April, and at Pili four months later, on 7 August. The battles are notable for their significant political consequences, as well as for the fact that the two commanders-in-chief were personally involved in the instigation of deeply flawed plans that led to humiliating disaster and significant loss of life. Both sides, therefore, have a continuing strong vested interest in drawing a veil over what happened during these battles, and why.

This analysis is based mainly on a study of readily available Maoist-produced videos, which cover both battles; and, in the case of Pili, by personal research that confirms and amplifies what is seen and heard on the videos. Tularam Pandey, a journalist with the Kathmandu Post, visited the Pili camp six days after the battle, and his reported accounts also tie in very closely with those from other sources.

This article first examines the battle at Khara. The PLA's first attack on the RNA base at Khara, in the mid-western Rukum District in May 2002, was repulsed with over 150 Maoists killed. It is hard to understand why the Maoist high command thought that it could succeed with another attack nearly three years later, knowing that the RNA had greatly strengthened the Khara fortifications. The video footage shows the base to be well sited on high ground, thus requiring any attacking force to fight uphill through minefields and elaborate barbed-wire obstacles, all capable of being covered by machine-gun fire. The fortifications also included a layout of well-prepared trenches and bunkers.

A conventional military assessment would have indicated that an attacking force would need a strong opening bombardment of artillery and mortars, perhaps supported by air strikes, to weaken the defences before assault forces could be launched with any chance of success. Even with such preparation, however, the attacking force itself would still need strong superiority of firepower in order to succeed. The Maoists enjoyed none of these advantages, and the lessons of previous failed attacks should have been clear to them. In November 2002, they had failed to overcome the strongly prepared RNA positions at Khalanga in Jumla District, a setback that the Maoists have recently acknowledged to have been a turning point in the war, and one that required a serious downscaling in their aspirations for overall military victory. In March 2004, the PLA likewise failed to overrun the RNA defences at Beni in west Nepal, and Khara was a much tougher objective.

Information on Khara is very difficult to come by. The battle has now been written out of Maoist history, and no member of the party seems prepared to talk about it. So why did this attack go ahead? Why the sensitivity and the collective amnesia about it? The answers are rooted in the bitter dispute between the Maoist chief, Pushpa Kamal Dahal (aka 'Prachanda'), and his longtime number two, Baburam Bhattarai, which came close to splitting the Maoist party in late 2004 and early 2005. Much is known about this feud because, as part of the reconciliation deal, the entirety of their acrimonious exchanges was made public, including being published on the Maoist website.

This dispute came to a head in January 2005, when a politburo meeting demoted Bhattarai, his wife Hisila Yami, herself a prominent Maoist activist, and a few other key supporters to the level of ordinary party membership. They also had restraints placed on their movement and outside communications. This dispute had a long history and many facets, but from early 2004 one key issue had begun to dominate: for some time, the Maoist leadership had known that there was no solely military way forward towards seizing state power. This truth had most likely dawned on some leaders as early as the setback at Khalanga two years earlier. All eventually came to agree that an alliance was needed, but the two sides remained divided over whether it was to be with Nepal's political parties, facilitated by India, or with then-King Gyanendra and his army. Bhattarai crystallised the division starkly when he characterised his opponents as "those who consider feudal autocracy as more progressive than capitalistic democracy."

We now know that, in late 2004, Prachanda was involved in direct talks with personal representatives of Gyanendra. There was the prospect of an imminent meeting between the two and even – or so it is alleged – the carrot being dangled before Prachanda of becoming prime minister. Then came the body blow of 1 February 2005, when Gyanendra launched his coup and seized absolute power. Journalist Bharat Dahal, in an August 2007 article in Nepal magazine, asserted that this action led directly to the decision to launch the Khara attack: 

The timing of action against Baburam coincided with the time when Gyanendra seized power. To Prachanda, all the doors to the palace, India and Baburam were closed. He prepared a draft of the attack on Khara in order to prove his 'brilliance'. The Party's huge armed forces were mobilised for the attack, but the plan failed miserably.

This allegation could hardly be more damning – that Prachanda chose to attack the very strongly held Khara position to show that, despite what Bhattarai and his supporters were saying and what Gyanendra had done, there was a military way forward to seizing state power. The evidence from the videos shows that the allegation is soundly based.

Prachanda's folly
Prachanda committed both the PLA's Western and Central Divisions to the Khara attack. In a briefing to troops on the march, Janardan Sharma (aka 'Prabhakar'), the commander of the Western Division, linked the need for the battle directly to Gyanendra's seizure of power, and spelled out its extraordinarily ambitious aim. "Gyanendra Shahi's coup is not his success," he stated, using the surname that seeks to reduce the king to a commoner.

It is the success of the People's War. This dramatic change in circumstances has posed a question for us. Should we let the battle we have fought so far collapse? Should we return to combatant warfare? Or should we move forward? Should we demolish our base areas and let all the people there be massacred and destroyed, or should two to four hundred of us die in order to open the door of final victory? The question lies here.

Opening the door to victory and conquest is a key motif repeated by all Maoist brigade commanders and commissars interviewed before the battle. Nanda Kishor Pun (aka 'Pasang'), the Central Division commander, was given overall command for the operation. In his orders, the provenance of the phrase was made crystal clear: "Let us take our revolution to that level, and let us really open the door to revolution, or in the words of the Chairman, let us 'open the door to conquest'. It is with that commitment that we have fixed this target." Another reply, by Bijaya K C (aka 'Bikalpa'), the Eight Brigade vice-commander, underscores the point: "Our Chairman has a dream, which is very much linked to reality. And that dream relates to conquest."

These commanders, like their peers, are full of confidence about victory in the battle, and about its aftermath. Birendra Budachettri (aka 'Jiwan'), the Third Brigade commander, asserted that "after previous battles, we got plenty of rest. But now we will launch attack after attack; we will achieve success after success. We will destroy the enemies' forts, and advance by sweeping away the enemy." The sentiments in Jiwan's words are typical of many. One can only conjecture how battle-hardened Maoist military commanders got themselves into such a delusional state of mind, but it is clear that Prachanda personally convinced them of the need for the battle – that it could and must be won, and that success would open the way to final victory.

The final orders for the attack were given using a very large and detailed model, built to represent Khara's features and defences. In the video, Pasang and Prabhakar are shown sitting side by side, and the body language between the two is not good. Prabhakar's contribution is muted, and a later interview with Bikalpa hints strongly at differences of opinion between the two divisions over the plan for the attack. To preserve the vital principle of unity of command, general military custom would have required the brigades of Prabhakar's division to be put under Pasang's command, with Prabhakar having no place in the command chain. In military operations, there must never be any doubt about exactly who is in charge; there can only be one overall commander and one aim, in order to ensure absolute unity and focus of effort. Any other arrangement risks confusion and disaster.

The video does not show scenes of actual fighting. But it does show elements of the lead assault groups being put into their final starting positions on the afternoon prior to the attack. What is striking is the scarcity of rifles. Many combatants had socket bombs, but many others were simply carrying picks, spades and wire-cutters, in order to dig under and cut through the barbed-wire obstacles. By 2005, the RNA had greatly strengthened the fortifications of all its bases, and there was no chance that a force as lightly armed as the PLA, however strong its fighting spirit, would succeed in overrunning the strongly fortified Khara position.

Some 250 were killed in the attack, the heaviest Maoist loss of the entire conflict. The attackers fought long and bravely against determined RNA resistance, but 'carnage' is probably the most appropriate word to describe what took place, as wave after wave of attackers were mowed down by machine-gun fire amidst the maze of barbed wire and strong fortifications. Indeed, with the odds so heavily stacked against them, it is extraordinary that the PLA was able to sustain the attack for more than 18 hours; most forces in a similar position would have cut and run much earlier. Nonetheless, low RNA losses indicate that, despite their persistence, the PLA was not able to achieve any significant penetration of the perimeter defences.

In an interview with Maoist journalists after the battle, Bikalpa complained that the battle ended before some of the people under his command had been committed to the attack. But his main criticism is more eye-catching: "There were many problems when we went to the battle," he said.

There were many weaknesses. As we analyse things, we had to bear consequences because of certain shortcomings. Things did not happen the way we had imagined they would because there was mischief and betrayal. Many things ended up deceiving us. Whose weakness was it the most? In clear and straight-forward terms, it was the commander's weakness. It was the main commander's weakness and also ours.

This criticism of Pasang is completely misplaced. The fact that it comes from a brigade commander in Prabhakar's division, and is filmed by a journalist attached to that division, simply confirms the suspicions expressed previously regarding rivalries and lack of unity of command. No commander of any army could have won the battle of Khara with the huge disparity in firepower faced by the PLA. Pasang would seem to have done well to finish the attack when he did: throwing in more troops would simply have added to the casualties.

This heavy defeat was a classic example of what happens when the strength of the moral component of military power is elevated and exalted as being supreme above all other considerations. Before World War I, the French and British generals committed the same mistake. But they learned the hard way, and at great human cost, that offensive spirit and high morale can only carry an attack so far when faced with the physical realities of barbed wire, well-prepared entrenchments, machine-gun fire and determined defenders. So it was for the Maoists at Khara.

For Prachanda, the defeat must have come as a crushing personal blow. His much-proclaimed dream about 'conquest' was shattered. This was indeed his battle, conceived and initiated by him for the personal reasons highlighted earlier. Afterwards, there was only one direction in which he could turn. Within four weeks of the Khara attack, Baburam Bhattarai, still technically reduced to being an ordinary party member, was in Delhi with one of Prachanda's right-hand men, K B Mahara, to start the process that ultimately led to the agreement with the Nepal political parties in November 2005.

The Khara defeat also left the Maoist military reputation bruised and battered. But within a few months, the 'feudal autocrat' in Kathmandu would himself order, for equally wrong and egotistical reasons, a deployment of his army. This would not only give the PLA the chance to regain its lost prestige, but also to acquire over 200 modern weapons and large amounts of ammunition and explosives.

Gyanendra's folly
On 24 July 2005, elements of an RNA combat engineer battalion began to deploy to build a camp on the steep-sided banks of the Tila River, a major tributary of the Karnali in the mid-western Kalikot District. No security-clearance operations were carried out in the area beforehand; nor was any action taken to provide adequate defences on the ground to cover the actual period of deployment and the building of the camp.

The decision was taken at the direct behest of Gyanendra. On his heavily guarded public-relations 'felicitations' around the country, he invariably would nod enthusiastically and affirmatively to endless requests for roads. Sadly for his pretensions as the great road builder, however, the general state of insecurity across the countryside made it very difficult to build anything in the rural areas. He saw his credibility as an effective ruler sinking by the day, as the Maoist insurgency spread and strengthened, and as all development activities stalled. To boost his public reputation, he decided that decisive action was needed. Thus, in early July 2005, at the height of the monsoon, he ordered his army chief to immediately resume the building of the Surkhet-to-Jumla road in Kalikot.

The likely consequences of obeying this command should have been spelled out to Gyanendra. Instead, the order was merely passed on, and the soldiers were dispatched to build a camp in what the RNA official spokesman, Brigadier General Deepak Gurung, later described with excessive candour as "a strategically unfavourable place". A second report, from 12 August 2005, quoted him as saying that "the temporary security base at Pili was not an ideal location. The decision to set up the base there was a technical one, not a tactical one. We didn't expect our workforce to bear such an attack." It was a massive dereliction of duty not to have done so. This failure also showed a complete disregard for one of the most basic tenets of war: never underestimate your enemy.

The Maoists heard about the RNA deployment on 2 August, 10 days after the first soldiers arrived at Pili. Prabhakar was with the Second Brigade at Dashera VDC in Jajarkot, 70 km southeast of Pili. Kali Bahadur Kham Magar (aka 'Vividh'), the vice-divisional commander, was with the Third and Eighth Brigades in Turmakhand village in Accham District, 70 km directly west of Pili but on the other side of the Karnali, preparing to carry out an attack on Martadi, the headquarters of Bajura District. (A VDC is Nepali terminology for a community cluster, called a Village Development Committee.) As soon as the PLA commanders heard about the Pili camp, they decided that it would be an easy target and should be attacked as soon as possible.

Initially, Prachanda was reluctant to give his approval. At the time, Maoist representatives were already in Delhi in consultation with the political parties, and a ceasefire was imminent. At the same time, Prachanda clearly could not risk going into the ceasefire on the back of a reversal on the scale of the failed Khara attack. He was assured by his commanders that success was guaranteed. The key commanders involved agreed that their forces should set off immediately with the aim of concentrating the division's three brigades at Raut hamlet of Pakha VDC, just one hour east of Pili, during the early afternoon of 7 August. The justification for the attack was well defined by Khadka Bahadur B K (aka 'Prakanda') the regional bureau in-charge, when he briefed commanders of the Second Brigade before they set off from Jajarkot: "In war, whichever side makes mistakes faces defeat. Our enemy has just made a big mistake, and we must move quickly to take advantage of it."

Getting to the objective on time required the three brigades to do a five-day forced march with little food on two different routes. The videos show that both routes were very tough going – at the height of the monsoon, and with a series of 4000-metre ridges that needed to be traversed. The Second Brigade travelled in an almost straight line from Jajarkot; the Third and Eighth Brigades first had to head northeast in order to cross the Karnali at the Jharkot bridge in Ramkot VDC, well to the north of Manma, the district headquarters. They then came south over the Chuli Himalaya to meet up with the Second Brigade at Raut village at 1 pm on 7 August. It was decided that they would attack that evening at six.

The attack achieved total surprise. The landing of an RNA helicopter 300 metres from the camp distracted the defenders, as they went to unload it. The Maoist command decided to take advantage of this opportunity, and the attack began at 5:45 pm. The Maoists quickly captured a number of unarmed RNA soldiers who had been unloading the helicopter, including the commanding officer of the battalion, who had arrived in the helicopter. The commander managed to escape during the subsequent confusion, and turned up two days later at Manma, along with 114 other men from the battalion who had also somehow managed to escape. Some evidence of their rapid exit and bedraggled state comes from Tularam Pandey's article. Along the trail from Manma to Pili, he recorded, "there were torn pieces of uniform, abandoned boots, caps and cartridges of bullets". He also quoted villagers as recalling how the soldiers who escaped were "hungry and naked", and that they gave the soldiers "food, clothes, shoes".

The RNA later announced that 227 soldiers had been in the camp at the time of the attack. There were probably a few more, since we know that 58 were killed, 60 were taken prisoner and 115 turned up at Manma. This means that about 120 soldiers stood their ground to resist the Maoists. More than 3000 Maoists took part in the attack, so it was not surprising that RNA resistance effectively ceased after two hours of fighting, by 7:45 pm. The hopelessly outnumbered RNA men, without a commanding officer and taken by surprise in what the video confirms was a very badly sited and unprepared camp, did very well to hold out for so long. The turning point in the battle came when the single medium machine gun in the camp, a GPMG, was overrun.

The Maoists were hungry and exhausted, so their first action after the battle was to cook a meal from captured RNA rations. Most slept near the camp, though some remained busy destroying and salvaging RNA kit and equipment during the night, an activity that may have misled some villagers about the actual duration of the attack. The first Maoists began withdrawing at six the next morning. By midday they had all left, having stripped the camp of all that was useful to them, and taking along 60 RNA prisoners. The first RNA troops arrived at Pili from Manma at 11:00 am on 9 August, a full day after the Maoist withdrawal. Twenty-six Maoists were killed in the attack. The RNA prisoners were released to the Red Cross five weeks later, on 14 September.

The propaganda battle
The aftermath of the Pili battle richly exemplifies the old adage that truth is the first casualty of war. Maoist journalists writing for Janadesh, the Maoist news outlet, with memories of Khara to expunge, had an obvious interest in exaggerating the length of RNA resistance and the strength of the defences. Elaborate, largely fictitious descriptions were offered of the attack. However, given the direct personal connection of Gyanendra and his army chief, General Pyar Jung Thapa, to the debacle, it was the RNA who had the greater interest in distracting people from asking questions about who was responsible. It quickly became apparent that there was to be no question of acknowledging mistakes or of paying proper tribute to the soldiers who fought bravely under such disadvantageous conditions. Instead, their sacrifice was to be derided and disparaged in an attempt to cover up for the gross incompetence and grave dereliction of duty of the top brass. Not even the dead bodies of the soldiers were to be exempt from being used in the public-relations exercise designed to achieve this nefarious purpose.

The first RNA effort at distraction was to say that that the great majority of those in the Pili camp were unarmed civilians. In a press conference held at the RNA headquarters on 12 August, this line shifted to make the contradictory claims that "the soldiers bravely resisted till morning" and that "most of the deceased soldiers were non-combatants who had not been provided with personal weapons." This story was repeated multiple times, despite the fact that combat engineers in all armies are trained as infantrymen in addition to learning their trade skills. At the district headquarters Beni in August 2004, it was an engineer battalion that had held off a force of over 3000 armed Maoists.

Another RNA claim was much more serious. On 11 August, Nepalnews, an online news organisation, stated that it had been given "exclusive access" to an RNA-produced video, which it claimed "revealed gory images of dead soldiers. The rebels had mutilated genitals, limbs and tongues of some of the soldiers before gunning them down, while some of the others were burnt alive with hands and legs tied tight with ropes." These were incredible claims to make on the basis of seeing a video. Without forensic expertise and personal examinations, no one can distinguish the nature of wounds resulting from a closely fought battle in which mortars and thousands of high-velocity rounds have been fired.

This claim was first made at an RNA press conference on 10 August, and was given wide publicity through the state-controlled media. The following day, the UN Office of the High Commissioner for Human Rights (OHCHR) urged the RNA "to ensure that a full forensic examination of the victims is conducted by qualified, independent experts". On 14 and 15 August, autopsies by such experts were carried out in Kathmandu on the bodies of the 58 RNA soldiers killed at Pili. The results were passed to the RNA, OHCHR and the National Human Rights Commission. The RNA immediately dropped the stories about mutilation and execution, presumably because there was no forensic evidence to justify them.

This was not the end of the obfuscation, however, and the next effort at distraction led the RNA to make its most ludicrous claim yet. On 13 August, the Kathmandu Post reported Brigadier General Dipak Gurung saying, at a press conference at the army headquarters the previous day, "Our soldiers have a major complaint that the INSAS [Indian National Small Arms System] rifles stop functioning after one to two hours of continuous firing. In Pili, the rifles could not function well after midnight. The soldiers resisted the Maoists till morning with the support of 81-mm mortar and GPMGs." This claim does not stand scrutiny. The Pili battle lasted just two hours, and a study of the relevant Maoist videos shows that most of the assault troops had a preponderance of socket bombs, and that the rifles they had were largely .303, with a sprinkling of captured INSASs and semi-automatic rifles. As a few observers commented at the time, the battle was fought with Indian weapons on both sides. The RNA soldiers were not disadvantaged by their weaponry – those who stayed in the camp to fight had plenty of excellent weapons.

A personal army
Writing 2500 years ago, Sun Tzu said that the apogee of strategic success is to achieve a situation in which an attack can be likened to "throwing rocks at eggs". Despite the arduous terrain and difficult conditions, the PLA showed great dash and skill to achieve at Pili a situation of advantage akin to Sun Tzu's ideal. However, they were greatly helped by the incompetence and extremely poor judgement of senior RNA commanders, who allowed their soldiers to be deployed to such a dangerous position without adequate protection. No effort at public relations can disguise for long a rout of the scale that took place at Pili – nor the monumental incompetence that caused it.

Those killed attacking the Khara base and defending the Pili camp join a long line of soldiers throughout history who have been victims of the ineptitude and egotism of their commanders. Some survivors of such disasters have been fortunate enough to see those responsible for their suffering and the fate of their comrades held to account and punished. There was never any chance of that happening in the autocratic, unquestioning world of Gyanendra's Nepal and the army he commanded. Nor was it ever likely to happen in the closed, secretive world of Nepal's Maoists.

With the return of democracy to the country, there is now the chance to ensure that, in the future, all those entrusted with the leadership of Nepal's security forces will be made fully accountable for their actions, and in a completely open and transparent way. However, with two armies still in existence, this will be a very difficult path on which to make much progress. The Comprehensive Peace Agreement between the Maoists and the government, which ended the conflict in November 2006, provided for a special committee to supervise, rehabilitate and integrate Maoist combatants, but the process stalled after only a single meeting. At the highest level, what is needed is a total transformation and streamlining of Nepal's bloated security sector as part of the progressive demilitarisation of what is, by any standards, a heavily militarised state. This is necessary both to save money for diversion to health, education and other programmes, and also to enable democracy to take stronger root.

The prerequisite to make any of this happen is to bring the present Nepal Army under firm political direction. Nepal's official army is now subject to less control from the state's governing authority than at any time in its entire history. In May 2006, instead of the previous palace-based authority over it passing to a functioning Defence Ministry or Prime Minister's Office, it passed directly to the army headquarters and into the personal hands of the current army chief. There is currently no sign of any political will to grip the generals, or to build the capacity to make civilian control of the military a reality – both essential foundations for a democratic state. The rarity of meaningful discussion on the subject is just one measure of the size of the task and of the moral courage required to champion its urgency and importance.

~ Sam Cowan is a retired British general who knows Nepal well.