AT about 10.30 am on a bright, sunny morning in early April, 1988, the city of Rawalpindi was rocked by a colossal explosion. Many people thought that India had attacked Pakistan or that our nuclear plant or bomb had been detonated. A massive, mushroom cloud of black smoke soared thousands of feet into the air. It heralded the start of a rain of rockets and missiles that continued throughout that day. The crash and crump of secondary blasts could be heard for the next two days. People 12 kilometres away were hit by falling rockets, although fortunately they were not fused, so were not exploding on impact. The entire arms and ammunition stock held by ISI at the Ojhri Camp for the Afghan war had gone up - all 10,000 tons of it. Some 30,000 rockets, thousands of mortar bombs, millions of rounds of small-arms ammunition, countless anti-tank mines, recoilless rifle ammunition and Stinger missiles were sucked into the most devastating and spectacular firework display that Pakistan is ever likely to see.

There was absolute chaos. One moment the roads around the camp were thronged with people, bicycles, carts and cars, the next the ground was littered with dead and dying. Almost 100 people died, and over 1000 were injured. These included five ISI staff killed and 20-30 wounded.

From the point of view of the enemies of the Jehad the timing was perfect. Within a few days the Soviets signed the Geneva Accord and the following month began their withdrawal in the knowledge that the Mujahideen had been deprived of all their reserves of ammunition at a stroke. The explosion occurred when the depot was fully stocked, in fact it was overflowing with at least four months' supply. While I was at Ojhri we tried to keep stock levels as low as possible, using the camp as a transit facility with daily deliveries forward to Peshawar. General Gul's new system of building up special composite packages of various types of weapons and ammunition for numerous Commanders required all kinds of supplies to accumulate at Ojhri. Only when sufficient quantities of all types had built up could individual arms packages be made up. As the CIA's part of the pipeline was so erratic, with shipments often being deficient of particular items, there was inevitably delay at Rawalpindi. On this occasion hundreds of packages had to lie at the depot for weeks. It was made worse by the fact that, for the previous three months, those packages destined to go straight to Commanders at the border had been held up because of the winter weather. At the exact moment when the Mujahideen would be expecting their depleted reserves to be replenished at the start of the spring operations there was nothing. If it was deliberate sabotage, it was a masterstroke. Before the last rocket had fallen the verbal accusations and recriminations were flying thick and fast. How could it happen? Why was so much ammunition stored in a densely populated area? Who was responsible? The civil authorities, led by the Prime Minister, blamed the Army and the ISI. The Army accused the ISI of gross incompetence, and both the Prime Minister and the Army turned on General Akhtar. It had been over a year since he had left ISI, but it was he who had authorized Ojhri as the main arms dump for the Jehad, so he must take the blame, or so his accusers thought. The fact that both the President and Prime Minister knew the camp's location, had visited it and had made no complaint as to its whereabouts were conveniently forgotten. This was the ideal opportunity to destroy General Akhtar and condemn the Army and ISI. President Zia, who was still Chief of Army Staff, was left with little option but to defend the military. His relations with the civil government and the Prime Minister were already strained, so there was no way he could meekly agree that it was all the Army's or ISI's fault. He supported his generals, Akhtar and Gul, in particular. The civil government used this tragedy to push their opposition to the military too far. Within a few weeks Zia had sacked the Prime Minister and dissolved the national and provincial assemblies when the Prime Minister tried to block the promotion of some generals, and insisted that the findings of any inquiry be made public.

An official military Court of Inquiry was immediately set up to investigate the disaster. It was headed by Lieutenant-General Imran Khan, the corps commander at Rawalpindi. He did not relish the job. There is no doubt in my mind that he did not know what to do, as he was being pushed by the Prime Minister in one direction and pulled by Zia in another. The former, I believe, wanted General Akhtar to be blamed, while the latter was insisting everything be hushed up, with no finger pointing at a culprit. The result was that Imran Khan dithered, which infuriated both his civil and military seniors. Eventually, perhaps not surprisingly, the court reached a finding that did not attribute blame to any individual. Whether or not the explosion was put down to an accident or sabotage I do not know for certain, as the court's conclusions were never made public. I do know that nobody was punished. Both Generals Akhtar and Gul continued in their careers. It was the Prime Minister who lost his job.

I was called as a witness to the inquiry, but was not greatly impressed with its methods or motives. Nevertheless, some basic facts emerged. A fire had started from one of the boxes containing Egyptian rockets, which had been sent to the ISI by the CIA for trials, before issue to the Mujahideen. Contrary to all safety regulations, these rockets had been armed with fuses by the Egyptians before shipment. A box fell down, either as a result of mishandling by the loading party, then in the warehouse, or due to a small explosive device. When it fell there was a minor explosion which started a fire. At that time several personnel in the warehouse were injured so there was a rush to treat and evacuate them by nearby staff. There was no attempt to extinguish the fire that had started as everybody was too busy moving the injured. After some eight to ten minutes the entire dump went up with one gigantic bang.

As to how it occurred there is no definitive answer that I know of. It could have been accidental, it could equally have been sabotage. If it was an accident then it could not have happened at a worse time as far as its effect on the prosecution of the war was concerned. The accident scenario has the fused Egyptian missiles falling due to mishandling; one went off causing a fire, perhaps in the wooden crate. The fire was not tackled as everybody was too concerned with the injured, so it took hold and set off the main explosion. This was not the first fire at Ojhri.

Almost exactly a year before fire had broken out in the same ammunition warehouse. On that occasion it had been due to some old World War 2 WP (white phosphorous) smoke grenades leaking and igniting. The NCO in charge had broken down the door and dragged out the offending box with complete disregard for his own safety. The fire was extinguished so there was no explosion. The inquiry recommended improved precautions. The staff at Ojhri were therefore conscious of the dangers and of the need to fight fires.

Those who feel it was sabotage base their argument on the fact that it could have been done and, perhaps equally importantly, on the perfection of the timing, on the amazing coincidence that at that moment the depot had never been so full, that the Soviets were about to start their withdrawal and they wanted to do so with the least possible harassment, and that the Mujahideen were depending on these supplies for their spring offensive. The sabotage theory has the rockets being tampered with in Egypt or in Pakistan, possibly at the request of the KGB. Once they were in the store then the device was detonated by a remote-control exploder from outside the camp. Alternatively, the device was planted by somebody who had access to the warehouse. It was guarded 24 flours a day, St! no outsider could enter. In this case the initial explosion? which caused the box to fall, could have been triggered by a timing or remote-control device.

If it was sabotage the Soviets had the most obvious motive, but. far-fetched though it may appear, the Americans also had reasons to wish the Soviets an uninterrupted retreat. As I have stressed, their policy was changing, they now wanted a stalemate, they wanted to prevent fundamentalists winning the war, and so Mujahideen without ammunition at this critical juncture coincided nicely with their objectives. The suspicion that, just perhaps, the US was not entirely blameless is heightened by the fact that the explosion was followed by the cutback in their shipments of arms. Had they really wanted to, I feel sure that strenuous efforts would have been made to replenish Ojhri Camp. No such efforts materialized; in fact it was not until the following December that further supplies arrived. The CIA knew that delivering arms at that time of the year effectively meant that nothing would reach the Mujahideen for a further three months, by which time the Soviets had gone. It all fell into place rather too neatly. For me, the destruction of all the Mujahideen's war reserves of weapons and ammunition was one of the turning points of the war. At the very time that the Soviets were pouring munitions and equipment into Afghanistan at an unprecedented rate, the Mujahideen were deprived of the means of carrying out any prolonged or large-scale operations. ISI remained in a state of shock for some time. They did not recover sufficiently to formulate strategic plans to clinch a victory in the field either during or after the Soviet withdrawal. A crucial period was wasted. To achieve the victory that everybody expected it was vital that the period of the withdrawal be used to plan, train, coordinate and dump the logistic requirements of the Mujahideen in various parts of Afghanistan. These activities should have been carried out in accordance with a sound military strategy, and should have been completed prior to the onset of winter. Nothing of the sort happened. No sooner had ISI begun to show signs of recovery than President Zia's aircraft was sabotaged, killing both him and General Akhtar. Within the space of sixteen months General Akhtar had been removed from ISI, Ojhri Camp was destroyed, the President, along with Akhtar and other senior generals, murdered, and the US was making it obvious that its support for the Jehad was now half-hearted at best.

In these circumstances the capture of Jalalabad was supposed to be the answer.

By early 1987 General Akhtar and I were confident, that it was only a matter of time before the Soviets quit Afghanistan. 1986 had witnessed Gorbachev's bleeding wound speech, their offer of a four-year withdrawal timetable, the actual withdrawal of six regiments and the introduction to the battlefield of the Stinger. We began to discuss an operational strategy to cover this event and to bring the war to a successful conclusion after they had gone. General Akhtar's relations with the Americans were somewhat cold and formal. He told me on several occasions that he did not trust the US to continue to support the Jehad wholeheartedly if the Soviets withdrew. I was inclined to agree, as I knew their antipathy towards fundamentalism and their desire for a moderate-nationalistic postwar government in Kabul.

One of the most difficult decisions that a guerrilla commander has to make once his forces begin to get the upper hand is the precise moment in the campaign when he should go on the offensive, when he should progress from guerrilla to conventional strategy and tactics. It is a matter of shrewd judgments. He has to assess the enemy's position with care. Is he sufficiently weakened numerically and materially? Is he demoralized, collapsing from within? Does he lack the means to keep his units adequately supplied? If the answer to these questions is yes, then perhaps the time is ripe to shift to the conventional phase of guerrilla war. But before doing so the commander must also examine his own forces. Are his men sufficiently trained to adopt coordinated conventional attacks, and if so on what scale? Are they well equipped with heavy support weapons? Can they cope with the enemy's likely control of the air? Can the scattered groups be supplied, concentrated, and then cooperate in joint offensives? Again if the answers are affirmative, then, probably, it is time to launch the offensive that will end the war.

There are numerous instances in military history when the guerrilla commander has moved into the conventional phase too soon, got a bloody nose, and as a result the campaign has been set back for months, even years. General Giap made this error in the early fifties against the French. The Communist Tet offensive in early 1968 in Vietnam failed, with losses of around 45,000 men, because their assaults were badly coordinated, communications were poor, the South Vietnamese Army fought well, and there was no demoralization within it ranks, or among the South Vietnamese population. Both the French and Americans lost eventually, but their opposing high commands had misjudged the timing of raising the stakes.

General Akhtar and I had considered this matter and decided that, even without the Soviet ground forces, it would be too risky to switch to a conventional strategy. Before General Akhtar was promoted away from ISI we had formulated an operational strategy to be applied during, and after, a Soviet withdrawal. Its objective was collapse in Kabul. If the people of Kabul, if the Afghan Army in Kabul, gave up, the war was won; but we did not feel this would be possible by direct assault. Kabul must be cut off, starved of food, fuel, men and munitions; the garrison must be demoralized and deprived of the means to fight. Then, and only then, were we confident they would surrender or turn on their Communist leaders. We did not consider the Mujahideen were ever likely to be ready for a conventional attack, or that one was necessary. We agreed that the strategy of a thousand cuts should continue, but with the emphasis on Kabul and its supply lines.

Map 22 shows what we had in mind. Kabul was to be surrounded by Mujahideen bases from which attacks were to be continuous. The Koh-i-Safi area would provide the main base for our efforts against Kabul which had to be made unusable. A series of blocking positions were to be established along all the main lines of communication from the Soviet-Afghan border to Kabul and Kandahar to deny logistic support to the Afghan regime. The strongest blocking positions were to be around the Slang Tunnel, the choke point for Kabul. We hoped that threats there would draw out forces from Kabul to clear the route, thus providing us with good ambush opportunities. Finally, the Mujahideen would contain and fix, as distinct from assault and capture, all the remaining Afghan garrisons in Afghanistan.

We could not of course decide on the timings for implementation as these would be dependent on the Soviets' withdrawal time frame, the weather (winter), and on our being able to bring forward our logistic requirements to the right places. Before we could take this strategy further General Akhtar left ISI and I retired in August, 1987. April, 1988, saw the Ojhri camp disaster; the Soviet withdrawal started the following month; President Zia and General Akhtar died in the air crash in August; the US cutback on arms supplies started; the 1988-89 winter was particularly severe. The Soviets had gone by mid-February, 1989, and in March the Mujahideen took to conventional warfare with a full-scale assault, not on Kabul, but on Jalalabad.

Why was such an attack mounted? Why was there no strategic plan to finish the war after the Soviets had gone? These questions are difficult to answer. Part of the problem was the euphoria, the elation, that gripped everybody at the prospect of imminent, easy victory. Certainly the Mujahideen Leaders and Commanders made the fatal mistake of assuming that the Communist government would collapse by the middle of 1989, that without the Soviets' presence its defeat was inevitable. This attitude was enhanced by the fact that most Mujahideen were busy making postwar plans and political maneuvering. The Afghan Interim Government (AIG) had been formed in December, 1988, and was sitting in Peshawar. Although unrecognized internationally, its members saw themselves as about to take over in Afghanistan within a matter of months. Peshawar politics became more important than military operations. There was a feeling that Peshawar was the place to be, securing a position in the AIG rather than asking life and limb in the field when the war was all but won.

The AIG, which was basically controlled by the seven Parties, backed be Pakistan, and seemingly with the support of ISI, selected Jalalabad IS their target of their post-Soviet strategy. It was to be a conventional attack on a major city[ but not the key city of Kabul]. The time had come. so they thought, to abandon guerrilla warfare. Jalalabad was tempting because it was so close (50 kilometres) to the Pakistani border of the Parrot's Beak. This meant that Mujahideen reinforcements and supplies should have quick and easy access to the front line. A main road led over the Khyber Pass to Peshawar. A victory at Jalalabad would enable the AIG to move forward with ease to Jalalabad. There they could declare a part of Afghanistan liberated and a new government established. This political objective had some merit, but depended for its fulfilment on military success. Could the Mujahideen surround and storm the city, and if they did would the Afghan Army collapse, or would they just get bombed out of existence? Above all would the loss of Jalalabad also lead to the loss of Kabul?

I believe General Gul allowed himself to be persuaded that it was militarily a sound proposal, partly by some of his younger operational staff, partly by the Leaders, and also by pressure from the Pakistan government, who saw it as a way of shifting all the Peshawar politicians and their countless followers back into Afghanistan. The easy capture of smaller garrisons at Barikot, Asmar and Asadabad in the Kunar Valley added to the Mujahideen's over-confidence.

In contrast to the dubious military strategy of the Mujahideen, that of the Afghans was simple and sound. To survive they had to hang on to Kabul and, if possible, the major population centres and military bases. Map 22 shows their strategic situation. To succeed in this they must concentrate their resources of men and munitions, and not be concerned if minor posts fell. They had to retain the ability to reinforce key positions by air if necessary, and above all they must keep Kabul supplied with food and military supplies. With logistic support the Soviets continued to be more than generous.

Although only a few advisers remained, Afghanistan was still the Soviets' war, as Vietnam remained an American one even after they too had left their allies to fend for themselves. Vast infusions of money and materials arrived. The war was able to continue due to the massive in-place transfers of weapons and equipment as well as the huge re-supply effort. In 1988 over 1,000 armored vehicles were handed over by the departing Soviets. It is estimated that the first six months of 1989 saw the transfer of $ 1.5 billion of military support to the Kabul regime, including 500 Scud surface-to-surface missiles. The Afghan Army still had tremendous superiority in what I call the three As - armour, artillery and aircraft. If they could bring these assets to the battle, if they could combine them effectively, then the Mujahideen would be defeated. The initiative was with the Mujahideen, hut they had to use it both strategically and tactically.

By March, 1989, the Mujahideen had assembled 5,()00-7,000 men in the hills around Jalalabad. However. their impending attack would not achieve surprise, as it had been heralded with too much publicity. The Jalalabad garrison knew what was coming and had made the necessary preparations. The 11th Division had been brought up to strength and other reinforcement units deployed in a ring of defenses. Bunkers, barbed wire and extensive minefields surrounded Jalalabad. The outer defenses extended 20 kilometres from the city, particularly to the east. Highway l, the link to Kabul, was protected by scores of posts throughout its length. Map 23 depicts the approximate layout of the Afghan defences and the main topographical features of tactical importance.

The Mujahideen assault began in early March with a direct, frontal attack from the east, up the Kabul river valley and on either side of Highway 1. Their first objective was the Samarkel position on the road 12 kilometres SE of Jalalabad. The ragtag warriors stormed ahead under cover of a heavy rocket, mortar and machine-gun barrage. Their initial impetus and enthusiasm carried them forward. The ridge east of Samarkel fell, and shortly afterwards the little village itself. Next the airfield, only 3 kilometres from the city, was taken by jubilant warriors yelling their war cries. This advance was led by several captured T-55 tanks crewed by the guerrillas. I believe this was the only tank versus tank engagement of the war. The Mujahideen's' success was short-lived, as the coordinated use of the three As drove them back from the strip.

The battle gradually became a stalemate, with more and more Mujahideen being sucked into the siege, but unable to coordinate their efforts, and wasting lives in reinforcing failure rather than success. Although some eight senior Commanders and their groups were deployed, there was no overall leader who could command obedience or devise a sound tactical plan. Attacks were invariably by day, with the Mujahideen walking or cycling forward in the early morning for a day's shooting, and returning at dusk to sleep in the deserted villages in the surrounding rich farmland.

From the outset a steady stream of miserable refugees, old men, women and children, tramped towards Pakistan. By June 20,000 had gone. Meanwhile the siege of Jalalabad ground on with the Mujahideen unable to improve on their initial success. A decisive factor in the attackers' failure was a lack of cooperation between the Commanders. They attacked when the mood took them, and without thought to concentrating or coordinating their efforts. As one exasperated Commander was quoted as saying in the London Sunday Times, 'There is no coordination. If the Mujahideen attack on one side and keep the government busy, the Mujahideen on the other side are sleeping'. This lack of an overall plan led to many setbacks. The vital highway to Kabul, which, after the airport was closed, was virtually the only way of reinforcing or supplying Jalalabad, was seized by the guerrillas. But instead of closing the road permanently, the Mujahideen kept rotating the groups responsible which enabled the enemy to keep slipping convoys through.

All through April, May and June the position of the Mujahideen gradually worsened. Within a matter of a few weeks ammunition shortages became critical. The heavy, and at times wasteful, expenditure in the early days could not be made good. The US shipments were still substantially less than necessary, the reserve stocks had never been built up again after the Ojhri Camp disaster, and there had been little forward planning or dumping of available stocks prior to the battle. Not only was the strategic wisdom of attacking Jalalabad doubtful, but the tactics and logistics of carrying it out were quickly revealed as inadequate.

The Afghan resistance had also been underestimated. These soldiers had to fight to survive. Some early killings by the Mujahideen of prisoners confirmed in their minds that surrender was no answer. They were supported by enormous firepower; they had the advantage of being in strong defensive positions; numerically they equaled, if not outnumbered, their attackers, and their logistic needs were met. Aircraft, including Antonov-12 transports converted into bombers, flew up to twenty sorties a day. Heavy bombs, and cluster bombs that exploded above the ground scattering scores of bomb lets over the target area, were used. These were extremely lethal against infantry. The effect on the ground is rather like the effect on a pond of throwing a fistful of gravel into the water, but over a far wider area. Although the Antonov is a slow-moving, propeller-driven aircraft, on these bombing missions it kept high, above the Stingers' ceiling.

Then there was the psychological as well as the physical effect of the Scud missiles. At least three firing batteries of these missiles had been deployed at Kabul, where they were maintained and operated by Soviet personnel. They were new weapons, introduced to help compensate for the Soviet troop withdrawal, and they were technically complicated, which explained the Soviet crews. A battery consisted of three launcher vehicles, three re-loading vehicles each with one missile, a mobile meteorological unit, a tanker vehicle towing a pump unit on a trailer, and several command and control trucks. Getting ready to fire took an hour. It involved a lengthy survey procedure at the firing position, using theodolites and optical devices, being completed before the missile could be raised upright for launching.     

The Scuds fired in Afghanistan carried high-explosive warheads weighing over 2,000 pounds. Jalalabad was comfortably within - range. The only warning the Mujahideen had was if they heard the sonic bang as the missile crashed through the sound barrier. They were area weapons. that is they could not achieve great accuracy. Their manuals indicate that when firing at a range such as from Kabul to Jalalabad, about half the missiles would fall in a circle with a radius of 900 meters. Over 400 Scud missiles thumped down among the hills around Jalalabad during the siege. I believe at least four fell inside Pakistan.

In four months of fighting the Mujahideen failed to take Jalalabad. It came as no surprise to me, or anybody else who took the trouble to study the situation. Their losses in men exceeded 3,000 killed and wounded. They expended what little reserves of ammunition had been accumulated, and their inability to breach the minefields and fixed defenses boosted the morale of their enemies. The battle for Jalalabad renewed the Afghan Army's confidence in its own ability, as well as telling the world that the Mujahideen were not yet able to march into Kabul. It was another major setback to the Jehad, from which the Mujahideen have not recovered to this day. Nor do I believe that their leadership has understood the lessons.

ISI and the Party Leaders made a strategic blunder in moving from guerrilla to full-scale conventional warfare too soon. They compounded it by selecting Jalalabad, whose capture would not necessarily bring down the Communist regime, instead of Kabul which would. They made no attempt to tie down Afghan reserves by keeping up the pressure at airfields such as Kabul or Bagram.

It was during the latter stages of the siege that Hekmatyar's men ambushed Massoud's forces in Takhar Province, sparking off the campaign of vengeance that resulted in the public executions described in chapter eight. That outright civil war should break out among the Mujahideen at such a critical juncture is indicative of the rapid erosion of what little unity was left for the Jehad.

Tactically, it was a textbook example of how not to fight a battle. There was no surprise; inferior forces attempted to assault prepared positions frontally in daylight. The attacks were poorly coordinated and the Mujahideen were subjected to a continuous barrage of shells and bombs from which there was no respite. Logistically it was grossly mismanaged. Due to the US cutback and the loss of all the strategic reserve stocks of arms at Ojhri, there was insufficient ammunition for a large-scale offensive lasting more than a week at most. The Mujahideen leadership knew all this, but still persisted with their plan.

General Gul was removed from his post at ISI in June 1989, when it was clear to everybody that Jalalabad was a catastrophe. His two-year involvement with the Jehad must have been a bitter experience for him. He came at a time when military victory was in sight; he left when Mujahideen defeat was distinctly possible. The falling away of American support, the Ojhri explosion, the air crash which killed the President, the fractious political infighting of the Leaders, which increased markedly as the Soviets left, and finally Jalalabad, demanded a scapegoat - General Gull Prime Minister Benazir Bhutto had him transferred back to the Army whence he came.

His replacement was General Shamsur Rahman Kallu. He was brought out of retirement for the job. Zia had got rid of him for having the temerity to suggest that the President should relinquish the post of Chief of Army Staff. He has closely followed the American line, bending to their pressures and thus effectively scuttling the chances of a Mujahideen victory. He has failed to retain unity among the AIG.

The Jehad has never recovered from Jalalabad. The Mujahideen had showed the world that they had the courage and skill to apply the pressures of guerrilla warfare to bring about the retreat of a superpower. Given the means to fight, given the cause of Jehad, and given a modicum of sensible military leadership, they could not be defeated. Take away these props and no army can win. Military history is a great teacher for both soldiers and politicians. Its lessons are few and of repeated. The problem lies in the learning.