Jammu & Kashmir: State Response to Insurgency - The Case of Jammu
Thomas A. Marks^*

Terrorism targets persons and property normally considered protected under the laws of war. Whether used as a strategic end – true terrorism, sociologically speaking – or as a weapons system within a larger insurgent campaign, terror confronts the state with the same challenge: how to create a security net – or grid – that negates the perpetrators’ ability to choose time and place.

Critical restraints are manpower and resources. Quantitative input, of course, is multiplied by qualitative factors. What results is that any insurgency or counter-insurgency must be assessed at different levels: macro, meso, micro. A case will only exist as an aggregate of its pieces.

Jammu and Kashmir (J&K, See map) illustrates well this dynamic. On the one hand, terror – as the driving force of insurgency by radical Islamic elements – appears to continue unabated in this, India’s northern-most State. On the other hand, disaggregating the case demonstrates that Delhi, after much experience, has evolved an effective response posture. This is particularly visible in Jammu, the southern division of the larger J&K.¹

Macro: Insurgency Considered as a Whole

At the time of Partition – the division of the provinces of British India into India and Pakistan – the additional 562 integral Princely States, one-third of the Indian land mass,² were called upon to join one or the other of independent Hindu-majority India or Muslim-majority Pakistan. J&K, one such princely state, had a peculiar problem, having a Hindu ruler but a majority Muslim population.

As the ruler (Maharaja) temporized, Pakistan endeavoured to force the issue with an invasion in October 1947, with the result that the panicked J&K Maharaja acceded to India.³ Pakistan has spent its entire independent existence struggling to reverse that result.⁴

Pakistan’s very identity has dictated this course of action. Formed as an Islamic nation in conscious opposition to India’s secular nature,⁵ Pakistan in a sense has been compelled to continue its quest for J&K, the only Muslim majority state in India.⁶ J&K’s position under Delhi’s sway challenges the basic reason for Partition – the claimed existence of ‘two nations’ within the larger subcontinent, each requiring a separate homeland. As a consequence, in each of the three major wars fought between the two states (1947-48, 1965, 1971), J&K has loomed large, with a portion, Pakistan occupied Kashmir (PoK), remaining alienated under Islamabad’s sway.⁷

‘Islamisation’ in Pakistan, the increasingly successful effort of powerful elements within the state to remould it along Islamic lines,⁸ has solidified the campaign to wrest J&K from India. Second only to the ‘holy war’ (jihad) in Afghanistan during the Soviet intervention (1979-89), the Kashmir campaign moved to a paramount position in Pakistani foreign policy after Moscow’s defeat.⁹

Events took a dramatic turn when, in the second half of the 1980s, missteps by India culminated in popular upheaval as a result of tampering in the 1987 state elections. Increasing militancy, centred in the Kashmir Valley, saw a temporary loss of Government authority, both State and Central. Led by the Jammu & Kashmir Liberation Front (JKLF), the internally generated insurgency demanded independence.¹⁰ Training, weapons, and equipment were increasingly secured in PoK, but the movement remained an internal phenomenon until Islamabad moved decisively from 1989 to support rival elements that sought not independence but union with Pakistan.¹¹

These groups proliferated rapidly – one Indian Army count detailed 177 different organizations.¹² A more recent count published in 2003 included 31 major groups.¹³

Necessarily, Pakistani involvement, conducted through the Inter-Services Intelligence (ISI), changed the nature of the struggle.¹⁴ Rather than being an insurgency per se, it became connected with what in American doctrine would be called an unconventional warfare campaign, the use of Pakistani intelligence (and special operations) personnel to train external jihadis to augment indigenous rebels. By 1999, more than half of those insurgents killed in J&K by the security forces were identified as foreigners (see Table 1).¹⁵

Terror, staged from camps in PoK, played a central role in the effort. Such actions did not remain limited to J&K, though. Increasingly, the ISI worked with disaffected Muslim elements in India – at 125 million in 1991,¹⁶ the country’s Muslim population was (and is) the world’s second largest, behind that of Indonesia – and with criminal syndicates, to wreak havoc. Bombay (now Mumbai), for instance, India’s financial centre, suffered major attacks in both 1993 and 2003.¹⁷

Just as Pakistan sought to enlarge the conflict to include India proper, so did it endeavour to ‘internationalize’ the struggle. That is, it constantly sought to engage in actions that would mobilize external pressure on India to engage in negotiations concerning the status of J&K. This principle has been a constant factor in guiding Islamabad’s use of conventional force to back unconventional action.¹⁸

In its most serious such gambit, Pakistan sought to take advantage of perceived politically favourable conditions within J&K, as well as the mutual possession of thermonuclear weapons by both Pakistan and India – which it saw as precluding a conventional counter-thrust by New Delhi – to launch a major unconventional operation in the Kargil sector of J&K.¹⁹ Regular Pakistani forces, disguised as insurgents, precipitated what rapidly became a full scale battle in May-June 1998.²⁰ A combination of Indian military pressure and the Pakistani decision not to become decisively engaged, resulted in a restoration of the status quo, but recriminations in Islamabad over the outcome were apparently central to the seizure of power by the then military chief, General Pervez Musharraf.²¹

Subsequent efforts to dampen tensions²² collapsed under the weight of two significant terror attacks: October 2001 on the J&K Assembly, in which 29 died; and December 2001 on the Indian Parliament itself in New Delhi.²³ Though the perpetrators were a combination of insurgents and jihadis, Pakistan was implicated in both assaults. With ‘9/11’ in the United States an intervening variable, the cumulative effect proved too much for New Delhi and it launched a general mobilization, Operation Parakram, for an apparent thrust into Pakistan.²⁴ Intense international pressure again proved decisive, but for Pakistan the post-9/11 environment, particularly the rout of its client Taliban in Afghanistan, brought a sea-change in world-context.²⁵

‘Internationalization’, Islamabad found, cut both ways, and the collapse of its Afghanistan position soon led to greater scrutiny of everything from its nuclear deals with rogue states to its support for J&K insurgents.²⁶ The latter had normally been given a free ride in the court of world public opinion, because of the ‘insurgent’ nature of the struggle. Such was an increasingly more difficult position to sustain, as jihadis became a majority of those killed. This trend accelerated as J&K popular support for armed militancy contracted, particularly in the wake of relatively clean State elections in September-October 2002, which seated the opposition People’s Democratic Party (PDP) as head of a coalition government.²⁷ The PDP’s efforts to find a peaceful solution to the situation meant little to the jihadis, who had become the driving force of the insurgency; and with their violence turning even upon their erstwhile supporter, Pakistan,²⁸ Indian forces faced the supreme irony of having dealt reasonably successfully with ‘insurgency’ while simultaneously witnessing a surge in violence against civilians in J&K – that is, in Kashmir itself.²⁹

Meso: Dividing up the Battle Space

There seems little doubt amongst observers that the situation in Kashmir is serious.³⁰ Yet this flies in the face of a finding that this is not so in all areas of J&K.³¹ What is needed, then, is to drop down a level from the macro to the meso.

Demographic and Physical Realities

Foreign jihadis do appear now to be the driving force behind the movement in all of J&K. They survive due to their use of terror and ability to maintain a support base. That there was invariably a supply of internal recruits stems from the geography and demographics of J&K.

Though official, popular, and academic treatments of the conflict speak of the J&K State as a whole – with statistics pertaining to the insurgency taken from the entire area – the situation on the ground requires clarification. A majority (69 per cent) of the J&K State land mass is ‘Ladakh’ (i.e. Leh and Kargil Districts), though this area has but 2.3 per cent of the 2001 census population of 10,069,917.³² Little if anything pertaining to the insurgency goes on there and it may be safely discarded in this treatment.

Insurgent activity, as indicated above, takes place in the divisions of Jammu and Kashmir, linked in an historic unity but in reality separated by a variety of factors – not least the Pir Panjal Range which runs between them – which dictate that even the rainy seasons occur at different months (for Jammu, peak rainfall is in July and August; for Kashmir, February-April). Jammu is larger physically than Kashmir (26,293 km² versus 15,948 km²) but slightly behind in population (4,395,712 versus 5,441,341, or 44.7 per cent versus 55.3 per cent).

Thus Jammu (26,293 km²) is slightly larger than either Vermont (23,958 km²) or New Hampshire (23,227 km²), with Kashmir (15,948 km²) being smaller than either but slightly larger than Massachusetts (12,549 km²). This comparison is important, because it means the entire conflict essentially goes on in an area (42,241 km²), just larger than the US tri-state block comprised of Massachusetts-Connecticut-Rhode Island (38,151 km²), but just smaller than the two-state block of Vermont and New Hampshire (47,185 km²).

In reality, the conflict, in terms of population, takes place in even more constricted space, with Kashmir’s population concentrated in the Vale of Kashmir (i.e. Kashmir Valley), and Jammu’s in the valley of that same name. Census data by decade reveals explosive population growth since independence, with the population essentially increasing by a third in each of the last two census-decades (i.e. 1981-1991 and 1991-2001). Indeed, given the 1951 census figure of 3.25 million, the present count of 10,069,917 makes for a 310 per cent increase in 50 years – with the greatest growth in the Kashmir Valley.³³

Jammu is dominated by Hindus (62 per cent), but three of its six districts have Muslim majorities (Poonch, Rajouri, and Doda; the other three districts, which have very large Hindu majorities, are Jammu, Kathua, and Udhampur). Kashmir’s six districts (Kupwara, Baramulla, Srinagar, Budgam, Pulwama, and Anantnag) all have Muslim majorities in excess of 90 per cent. Hindus, in fact, were reported to be less than 2 per cent in all districts of the division except Srinagar, where their numbers were placed at 6-8 per cent.

Since the State as a whole (certainly Kashmir Division) remains tied to the employment patterns generated by agriculture, all sources have noted post-independence employment problems, especially the high dependency ratio (i.e. the number of persons supported by the working population). In 1981, less than a third of the State population (30.4 per cent) was classified in the census as ‘main workers’ (i.e. those working more than 6 months, 183 days), with another 13.9 per cent classified as ‘marginal workers’ (finding less than 6 months employment), and fully 55.7 per cent as non-workers. As might be expected, those in age group 15-39 dominated the labour force, and nearly three-quarters (72.0 per cent) of those working were in occupations tied directly to primary activities, and to the land (e.g., cultivators, agricultural labourers, livestock workers).

At least two decades ago, then, issues of livelihood for the young had been identified as a looming State problem, with all factors exacerbated in Kashmir by Islamic cultural traits (such as discrimination against women and preference for male offspring). Already, in 1981, more than half the State population was less than 19 years of age, with a literacy rate well below the national norm (and even lower among Muslims and especially Muslim women). Significantly, the lowest level of agricultural employment in the State was in Srinagar District (16.7 per cent), which was tied to small shop-keeping and thus dependent upon external forces for generation of employment capacity.

As this heavily Muslim district was also an area of explosive population growth, the ability of the economy to absorb youth steadily declined – even as the cultural bias noted above produced a pronounced imbalance in the sex ratio (for Kashmir the Female: Male ratio in 1981 was 878:1,000 versus Jammu’s 925:1,000). Population density was considerably higher in Kashmir than elsewhere in the State, 251/km² as early as 1991 versus 135/km² in Jammu (and just 2/km² in Ladakh).

The upshot is a statistical case can be made that there was a demographic tidal wave of unabsorbed youthful males appearing in the late 1980s, especially in Kashmir, just as political issues discussed above called into question the legitimacy of the existing order. Yet the resulting insurgency, despite its widespread violence in both the Jammu and Kashmir Divisions, is in its origins and driving force more a Kashmir than a Jammu problem. Indeed, the increasingly Islamic nature of the insurgents and their support from Pakistan has served to enflame latent separatist sentiment on the part of Jammu. One now sees strong forces demanding independent consideration of Jammu in factors ranging from political to linguistic; and local defence forces (to be considered below) in Jammu, at least, are dominated by Hindus.³⁴

Be all this as it may, a point must be made as concerns the earlier discussion of the need to examine the insurgency in its parts: it is not the human cost alone that makes for the notoriety of the conflict. Indeed, the internal war in J&K, when scaled, does not begin to approach the levels of criminal violence present in those U.S. metropolitan areas best known for their murder rates. The ‘death count’ in Jammu & Kashmir for 2003 stood at 836 civilians, 1,447 militants, and 380 security personnel.³⁵ If this violence is aggregated (2,663), which is unorthodox but certainly presents the worst possible statistical picture, it scales out at 24.5:100,000 population.³⁶ This would place J&K between Memphis (24.7:100,000) and Chicago (22.2:100,000), in the 2002 murder rankings when examining American cities with populations greater than 500,000, well off the pace established by the likes of Washington, DC (45.8:100,000) or Detroit (42.0:100,000).³⁷

Thus the issue, as concerns Indians, is not ‘body count’ alone but the totality of the dislocation. The perversion of daily life caused by the insurgency and the Government’s response; the deployment to the State of substantial numbers of security forces; the inability of economic activity to respond to demographic shifts due to the all-encompassing and pervasive effect of the conflict; the looming danger of escalation to inter-state war, with the possibility that nuclear weapons will be used; these and other facets are what make the Kashmir conflict so ominous for the population and for the country.³⁸

Indian Concept of Counterinsurgency

‘Original causes’ invariably become less salient as an insurgency progresses. Even as India’s response has matured and evolved – displaying more nuance and professionalism – so has the insurgency. The result is that what India presently faces has as much the character of a Pakistani special operation exploiting an unstable internal state situation than it does an internally-generated insurgency.³⁹ This does not cause ‘hearts and minds’ measures to be cast aside, but it does mean they must be accompanied by particularly robust population and resources control, as well as military measures.

This is not misguided, for it is doubtful that ‘original causes’ could sustain anything approaching the present level of insurgent activity absent the PoK sanctuary and the Pakistani provision of arms, ammunition, and equipment. Even now, despite the substantial numbers of Indian forces deployed and resources being expended, the conflict resembles more what was seen in Ulster (Northern Ireland) than it does mass-based illustrations (e.g. Nepal or Peru). Vocal, elite-driven support that continues to exist in Kashmir, for either independence or unification with Pakistan, can not be said to be shared by a substantial proportion of the Kashmiri population – and certainly has no significant backing within Jammu (or Ladakh).

What now exists, then, we see in not only Ulster but also the likes of the Basque country in Spain or Corsica in France. What distinguishes the J&K conflict from those cases is the role of Pakistan amidst the sufficient pool of recruits thrown up by State realities (which, to reiterate, is especially pronounced in Kashmir). This allows a militant movement to recruit and sustain operations even as it increasingly drifts from its purported mass base.⁴⁰

Faced with militancy, however defined, India’s response has been consistent and driven by a ‘support to civil authority’ doctrine. This approach stems from its adoption at Independence from the former colonial power, Britain. Insurgency therefore remains operationally, first and foremost, an issue of law and order, and thus is to be met by reinforcing the normal mechanisms of the state, most especially its local security forces (i.e. police). Though the military was interjected into the counterinsurgent dynamic in early 1990s and is clearly the dominant force in terms of sheer power, police primacy is the template within which all force dispositions take place.⁴¹

This has meant in particular the establishment of a legal framework for carrying out counterinsurgency. A variety of national and State ordinances have been implemented, allowed to lapse, then been resurrected. The most prominent have been the Terrorist and Disruptive Activities (Prevention) Act 1987, or TADA, and later the Prevention of Terrorism Ordinance, 2001, or POTO, later the Act of 2002 (POTA).⁴² Special procedures, special courts, special measures for protection, all are stipulated. The result is that while armed response by the state is carried out in a fashion associated globally with ‘emergency legislation’, it remains subject to control by and challenge through the legal system.⁴³ Just as importantly, the framework of elected Government has remained unchallenged, though at times suspended.

For administrative purposes, J&K is divided into 14 districts (invariably referred to by security officials as "revenue districts"). These are the basic framework for the police district structure, but a number of administrative districts have been further divided into several police districts to improve command and control. Thus there are 21 police districts administered by six headquarters, called ‘ranges’ (see Table 2 below), three each for Jammu and Kashmir ‘zones’. Each zone is headed by its own IGP (Inspector General of Police) answering to the J&K Director General of Police (DGP), each range by its own Deputy Inspector General of Police. Each of the police districts is headed by a Senior Superintendent of Police (SSP), assisted by various staff (e.g. ranks of Superintendent of Police (SP) and Additional Superintendent of Police (ASP). Beneath the districts are any number of police stations headed by various ranks depending upon size and situation. These may be ASPs or Inspectors. In the police complement, there may be any number of Inspectors, Assistant Inspectors, and Sub-Inspectors.

Police stations vary in number per police district but are not particularly numerous, ranging from 6 to 20. Poonch police district (coinciding with Poonch revenue district), for instance, has just six stations and 3 ‘police posts’ for its area of 1,674 km² and population of 371,561 (Census 2001).⁴⁴ Different numbers of personnel are assigned to each station and post, of course; but again, the normal number of officers is low considering the populations administered. Mandi police station (for all locations in this paragraph, see Map 2 below), for instance, 23 km from Poonch and under that police district, oversees approximately 60,000 people with just 38 regular officers and 45 special police officers (SPO; to be considered below).⁴⁵ Loran police station, also under Poonch police district, has 25 regular police officers and an equal number of SPO to administer to "5 villages containing about 20,000 people."⁴⁶ Police post Sawjian, under Mandi police station, has 11 regular police officers and 32 SPO to attend to the needs of "4 villages with about 17,000 people."⁴⁷

Map 2: Poonch (Tourist Attractions Handout)

These figures are in keeping with those found in other developing nations, such as Colombia or the Philippines. Police force strength for the entire State is 60,000 (of whom just 2,000 are women).⁴⁸ In normal times, the Police complement is totally comprised of regular police officers. Their activities are as would be expected, with all interviewees reporting their primary criminal concern as ‘trespass’ (i.e. occupying someone else’s property, which in practice normally means land). Violent crime appears to be a minor factor in all jurisdictions, as confirmed by examination of ‘incident boards’ in stations visited.

In abnormal times, though, India has a variety of ‘surge’ mechanisms for augmenting regular police strength. These are much in evidence in J&K and are the heart of the counterinsurgency programme. They combine State and national programmes as dictated by the Indian Constitution:

• A special category of VDC, which guards specifically the high pasture areas (Dhok) are constituted as Dhok Defence Committees (DDC). They function by the same mechanisms as the VDC but are distinct.
• The innumerable point-defence duties that are part and parcel of population and resources control, especially guarding key facilities and personalities, are often assigned to the functional equivalent of police paramilitary units, in this case the Jammu & Kashmir Armed Police (JKAP). Comprised of battalions, they are normally deployed in urban areas.
• To respond to the needs of police stations for area domination beyond the reach of their minimal resources (even as augmented by SPO and VDC), the Central Government, under agreement with the State, may deploy paramilitary battalions from the Central Reserve Police Force (CRPF). Raised in 1939 to aid State Governments in the maintenance of law and order, as well as internal security, CRPF had 137 battalions nationwide, comprised of 167,322 personnel by 1999 (with additional battalions being formed).⁴⁹
• Similar to the CRPF units are the battalions of the Indian Reserve Police Force (IRPF). Though raised and financed by the Federal Government, they are recruited and deployed locally, responding directly to the State Government. Though essentially clones, IRPF battalions are preferred to their CRPF counterparts not only because they are State assets but also because they have a more robust officer complement. J&K has nine of these, with two more forming.⁵⁰
• Providing additional security to infrastructure, especially Government facilities, is the Central Industrial Security Force (CISF), formed in 1969. Made up of 99,863 personnel (1999 figure), it is structured along military lines and is reportedly deployed to guard more than 230 public undertakings nationwide. It may work with private security firms, its actual deployment posture depending upon the enterprise to be guarded.
• Guarding stretches of the border is the Border Security Force (BSF), which was formed in 1965 to replace the multiplicity of state forces that were doing the same chore. At 157 battalions (1999 figure), with 176,000 personnel, it is a paramilitary force that has organic capabilities normally associated with the military, from air and water assets to artillery groups. It is primarily responsible for the India-Pakistan and India-Bangladesh borders, but it also has forces on the borders with China and Burma (the border with Nepal is open). Its capabilities have meant that it has been used often in ‘fire brigade fashion’ to respond immediately to internal security crises.
• Performing in a fashion similar to the BSF, but formed in October 1962 and charged specifically with securing the China border is the Indo-Tibetan Border Police (ITBP). Its 29 battalions and 30,367 personnel (1999 figure) augment BSF skills with expertise in counter-insurgency, counter-terrorism, and mountain warfare. Units are intended to be self-sufficient due to the harsh terrain in which they operate. These characteristics have led to their employment in stability operations as in the case of the BSF.
• This may also be said of the Assam Rifles, which traces its history to 1835 and duties to securing the Northeast frontier. With 31 battalions and 35,000 personnel (1999 figure), the entire force is under Army operational command (OPCON), conducting counterinsurgency in India’s Northeast. Characteristics, though, have meant units and personnel have been deployed elsewhere as dictated by emergency circumstances.
• If, after resort to all of these levels of police and paramilitary reinforcement, the situation still requires coercive power, the military will be committed, as it has been for a decade in J&K. The principal force is the Army, which has a total infantry strength of 355 battalions (1999 figure). In addition to these regular units, however, a decision was made in 1990 to form a special ‘counterinsurgency force’, the Rashtriya Rifles (RR), for the dual purpose of securing the rear area during conventional operations and engaging in stability operations during peacetime. In 1999, the force was already at 48 battalions (of 60 authorized), but steps were underway to stand up to authorized strength.

The intent is to provide layer upon layer of security forces in response to internal upheaval – a grid. Always, though, the point of reference is existing society, with emergency manpower, such as recruited mainly in SPO and VCD/DDC fashion, explicitly labelled as short-term and funded as such (pay in many cases more of a stipend nature than actual wages). The proliferation of paramilitary forces over the past decades has been directed at providing an emergency response capability that nevertheless remains within the civil authority structure and does not distort the day-to-day realities of law enforcement.⁵¹

Deployment of Forces: Jammu

Under Indian federalism, State police forces are independent. J&K’s forces are especially so due to the unique, semi-autonomous status afforded to the State by the Indian Constitution.⁵² The initial response to militancy was thus a State matter. With the passage of time, the security mechanisms became more linked to those of ‘the Centre’ (New Delhi) and came more to resemble those one would find in any such crisis in India.

In any insurgent situation, regardless of impetus (internal or external), the issue faced by the security forces is domination of human terrain. The low police/population ratio in J&K made this particularly difficult. Population was spread unevenly, with myriad detached settlements and even homesteads.

The manner in which this dilemma has been dealt with is revealed by taking the template above and applying it to Jammu (see Table 3). The key is to deploy interlocking layers – a grid – of both general and mission-specific forces, but all serving to reinforce the existing civil structure with its law and order component.

Micro: District Doda of Jammu

This can be made even clearer by examining a single district, Doda (see Table 4). In normal times, as indicated above, the number of police officers, though thin, 686, was considered adequate, given the general lack of criminal activity upon the part of the populace, 690,474. Faced with a systematic campaign of terror directed against the people and Government apparatus of the district, though, augmentation became necessary.

As a result, this single district was given authority to mobilize 7,400 SPO’s. While a normal police officer begins at a base monthly salary of Rupees 5,000 (US $111) plus allowances, an SPO earns just Rupees 1,500 (US $33) plus allowances on a contract basis. SPO’s, though invested with the same power and authority as civil service police officers, receive accelerated training and are deployed principally for operational missions, such as point defence and VIP security.

In Doda, this breaks down to 4,496 ‘operational’ SPOs. The remainder, 2,904 SPOs, serve an equally crucial function, providing the C2 personnel for the local defence forces (VDC and DDC). These number a substantial 9,545 personnel, with 8,999 of them armed with the .303 Lee Enfield rifles, a dated 1941 bolt action piece, but one still highly lethal in a defensive posture.

Not only do the VDC/DDC-assigned SPOs perform C2 functions, they also have automatic weapons capability, since they carry SLR’s or other high-powered firearms. As the militants generally operate in sub-section (i.e. sub-squad) strength, the VDC/DDC’s are more than capable of holding their own, pending reinforcement.

For their part, VDC/DDC members are citizens engaged in defence of hearth and home. Often they are paid as little as Rupees 400 (US $8.89) per month, but just as often nothing at all. Funds are in as short supply as weapons (thus the lack of arms for all).

Ironically, those initially most threatened, the Hindus, flocked to the VDCs/DDCs and thus came to dominate the units, exhausting funding and arms supplies, even as militant terror so traumatized the general populace that Muslims, too, asked for a self-defence capacity. This they have been given, but requests for additional means remain pending.

Regardless, what leaps out from the statistics above is that a single district, faced with a security crisis, can surge from fewer than 700 law enforcement personnel to nearly 18,000, virtually all armed – and all under the C2 and legal authority of the constitutional structure.

With such a front-line defence, the augmentation required by paramilitary and military forces (again refer to Table 3) is not overly large but is nevertheless potent: some 11 battalions versed in counterinsurgency techniques and armed more heavily than the police themselves. As the police focus upon the populated areas, the reinforcement forces seek to go after militant base areas and mobility corridors.

Only one capability in the tables above has not been specifically noted, that contained in the Special Operations Group (SOG). These are pseudo-gangs comprised of former militants who have rallied to the Government side. Normally, their C2 comes from the police. They have been quite controversial but very effective, perhaps more so in intelligence generation than actual combat. Whenever political debate has threatened their existence, they have been absorbed into the Army special operations structure and have continued to perform their missions.

What remains to be examined is the cumulative impact of this surge. Journalistic sources are wont to paint the worst possible picture.⁵³ Ground reality looks somewhat different: not ‘normal’ by any means but certainly ‘under control’. Peak figures available for the year 2001 show 51 Hindu and 66 Muslim civilians killed in the district; 35 security force personnel; and 101 insurgents.⁵⁴ Scaled, the 253 total deaths work out to 36.6:100,000.⁵⁵ This is more dangerous than J&K as a whole, less dangerous than the Kashmir Valley. It would be considered a serious problem in any society but is by no means extraordinary.⁵⁶

Conclusions

Therein lies the point. The insurgency in ‘Kashmir’ can not be considered as such. Neither can the counterinsurgency. Each of the pieces must be considered separately before any composite can be reached.

Likewise, assessing the state of the insurgency requires qualitative judgments that incorporate quantitative measures but do not overemphasize them. Counterinsurgency is frequently presented, by both supporters and critics, as an approach that either does or does not (respectively) ‘work’.⁵⁷ Reality is quite different: an approach that is both correct (i.e. addressing the root causes of the insurgency) and sustainable is put in place – and the state then ‘plays for the breaks’. Changes in ‘the situation’ may take a quarter of a century, as they did in Northern Ireland.⁵⁸

The situation becomes more difficult when, as is the case in J&K, the insurgents adopt terror as strategy rather than mass mobilization. That is, they seek, as was the case with the Provisional Irish Republican Army (PIRA) in Northern Ireland, merely to inflict as much ‘pain’ upon the state as possible to force its capitulation or simple withdrawal from the contest. When such a campaign is linked to suicidal action and external sanctuary, the Indian approach is the most viable possible.⁵⁹

The forces above are deployed in a ‘grid’ that seeks to dominate all areas. By fencing the entire LoC and engaging in active patrolling, the security forces endeavour to seal off the battle space.⁶⁰ The police work with the population; the paramilitary and military forces do the heavy-lifting in areas of minimal population and rough terrain. A recent Army operation, for instance, used an entire brigade to eliminate a substantial insurgent base area in the Hill Kaka area of District Poonch.⁶¹

To integrate such a substantial array of forces requires extensive coordination, but this is achieved principally through an understanding of the commander’s concept and regular use of land-line telephone. Formal C2 meetings, while they occur at regular intervals and involve all forces, as well as civilian representatives as appropriate, are kept to a minimum.

Perhaps of greatest moment, democratic authorities continue to have the ultimate say in matters of security. Elections, though at times disrupted for reasons of security – more often held in staggered format – nevertheless occur. This serves to put legitimacy on the side of the authorities and to give them the ultimate trump card in the battle for the hearts and minds of the afflicted citizens.

Those hearts and minds, to be clear, generally support the Government, a trend that has accelerated as terror has been increasingly used against the population rather than the security forces. What ‘the grid’ allows the Government to do is both to secure the populace and to wear down the militants.

This is critical, but in the changing environment it can not yet be decisive. India may, indeed, recognize that politics is at the heart of the conflict, but the foe has continued to evolve. Most significantly, the principal insurgent group, the Lashkar-e-Toiba (LeT) has made the transition to terrorism and increasingly extended its operations to India proper.⁶² Of most concern has been the involvement of disaffected Muslim individuals in terror actions, but this has thus far been limited and has not substantially widened the J&K conflict.⁶³

A diplomatic breakthrough remains a possibility, even if remote. In particular, Pakistan has continued to adopt a more nuanced position than was previously the case. The danger posed by the jihadis to the Pakistani polity itself has led to an effort to rein them in. Outraged, they have responded by at least twice attempting to assassinate President Musharraf, as well as other members of his inner circle.⁶⁴ Musharraf’s approach, though not particularly sophisticated or forceful, nevertheless recognizes full well that the lay of the geo-strategic land has definitely been altered.

As a consequence, though violence in J&K continues, key ‘moderate’ groups among the militants have met face-to-face with India’s Prime Minister.⁶⁵ Such progress would not have been possible had the security forces not held the line. Ironically, determined to scuttle any moves towards peace, the jihad has moved periodically to up the ante. Hence the relative escalation in violence against civilians. This, however, may mark as much a turning point in the insurgency as any hitherto seen.

• Dr. Thomas A. Marks is Professor of Insurgency, Terrorism and Counter-Terrorism at the School of National Security Executive Education (SNSEE) of the National Defence University (NDU) in Washington, D.C. This paper was completed during the four years he spent as an independent consultant prior to joining NDU

1. My most recent research visit to the Indian State of Jammu and Kashmir took place in August-September 2003. Though previous trips were made to Kashmir and Ladakh, this trip focused completely on the Jammu Division due to access afforded by unique circumstances. Readily available overviews of the conflict are: Victoria Schofield, Kashmir in Conflict: India, Pakistan and the Unending War, 2^nd ed., New York: I.B. Tauris, 2003; Sumantra Bose, Kashmir: Roots of Conflict, Paths to Peace, Cambridge, MA: Harvard University Press, 2003; and Sumit Ganguly, The Crisis in Kashmir: Portents of War, Hopes of Peace, New York: Cambridge University Press, 1997.
2. Useful as a single reference for the era of British rule and its administrative particulars is Geoffrey Moorhouse, India Britannica: A Vivid Introduction to History of British India, Chicago: Academy Chicago, 2000. As Moorhouse notes (p. 104), these "never came under the direct rule by the British. Amicable arrangements were made to ensure that these private enclaves did not obstruct the vital workings of the Raj: the British Posts and Telegraphs Department operated throughout the princedoms and the railways crossed them without interruption. In exchange for such concessions, the princes were allowed to raise their own military forces…"

3. For background, among myriad available titles, see Prem Shankar Jha, The Origins of a Dispute: Kashmir 1947, Delhi: Oxford University Press, 2003; Iffat Malik, Kashmir: Ethnic Conflict International Dispute, New York: Oxford University Press, 2002; R.N. Kaul, The Wail of Kashmir: In Quest of Peace, Delhi: Sterling, 1999; C. Dasgupta, War and Diplomacy in Kashmir 1947-48, Delhi: Sage Publications, 2002; S.P. Vaid, How Partition Rocked Jammu and Kashmir, Jammu: Shyama, 2002. For the conflict itself, see Kuldip Singh Bajwa, Jammu and Kashmir War (1947-48): Political and Military Perspective, Delhi: Har-Anand, 2003, and Agha Humayun Amin, "Grand Slam – A Battle of Lost Opportunities," Defence Journal, September 2000, http://www.defencejournal.com/2000/sept/grand-slam.htm.

4. For overview, Sumit Ganguly, Conflict Unending: India-Pakistan Tensions 1947, Delhi: Oxford University Press, 2002, and Robert G. Wirsing, India, Pakistan, and the Kashmir Dispute: On Regional Conflict and its Resolution, New York: St. Martin’s Press, 1998.
5. Among those central to Pakistan’s formation, there was no unanimity concerning the shape the new state should take or the course it should follow. Nevertheless, there was no division on the point that Pakistan was to exist as a homeland of sorts for India’s Muslims. See Mushirul Hasan, Legacy of a Divided Nation: India’s Muslims Since Independence, Boulder, CO: Westview Press, 1997; J.K. Chopra, Pakistan as an Islamic State, Jaipur: Sublime, 2000; and Akbar S. Ahmed, Jinnah, Pakistan and Islamic Identity: The Search for Saladin, New York: Routledge, 1997.

6. Background in M. Ishaq Khan, Kashmir’s Transition to Islam: The Role of Muslim Rishis, Delhi: Manohar, 1997.

7. Excellent, as a single source, is Navnita Chadha Behera, State, Identity and Violence: Jammu, Kashmir and Ladakh, Delhi: Manohar, 2000. This may be usefully supplemented by Owen Bennett Jones, Pakistan: Eye of the Storm, New Haven, CT: Yale University Press, 2002, esp. Ch. 3, "Kashmir," pp. 56-108; and Christophe Jaffrelot, ed., Pakistan: Nationalism Without a Nation? Delhi: Manohar, 2002. As detailed on the map, an eastern portion of the historic J&K remains under Chinese control as a consequence of the 1962 Sino-Indian War.
8. See e.g. Veena Kukreja, Contemporary Pakistan: Political Processes, Conflicts and Crises, Delhi: Sage, 2003, esp. Ch. 5, "The Rising Tide of Islamic Fundamentalism in Pakistan," pp. 154-90; as well as Pooja Joshi, Jamaat-i-Islami: The Catalyst of Islamization in Pakistan, Delhi: Kalinga, 2003, and Frederic Grare, Political Islam in the Indian Subcontinent: The Jamaat-i-Islami, Delhi: Manohar, 2001. Revealing, too, is Yvette Claire Rosser, Islamisation of Pakistani Social Sciences Textbooks, Delhi: Rupa, 2003.

9. Excellent as a single source is Barnett R. Rubin, The Fragmentation of Afghanistan: State Formation and Collapse in the International System, 2^nd ed., New Haven, CT: Yale University Press, 2002.

10. A succinct recounting can be found in G.M. Shah, "Geo-Political and Socio-Economic Causes of Militancy in Jammu and Kashmir," in V.D. Chopra, ed., Global Challenge of Terrorism, Delhi: Gyan, 2002, 189-98. Also useful is Peer Giyas Ud-Din, Understanding the Kashmiri Insurgency, Jammu: Jay Kay Book House, 1992. For the 1987 elections and the events surrounding them, see G.N. Aali, Restive Kashmir, Srinagar: Itme-Non, 2002.

11. See Behera, pp. 164-214. For a discussion of the important Rajouri-Poonch area, see Sudhir S. Bloeria, Pakistan’s Insurgency Versus India’s Security: Tackling Militancy in Kashmir, Delhi: Manas, 2000. The author, it should be noted, from his position as Special Commissioner and Special DIG of Police for the districts, went on to become the Principal Secretary to the Governor of J&K.

12. Ibid, p. 182; Appendix 6 (pp. 326-28) lists all groups.
13. K. Santhanam, Sreedhar, Sudhir Saxena, and Manish, Jihadis in Jammu and Kashmir, Delhi: Sage, 2003. Further discussion of major groups may be found in N.C. Asthana and Anjali Nirmal, Terrorism, Insurgencies and Counterinsurgency Operations, Jaipur: Pointer, 2001, 84-88. Analysis in Shekhar Gupta and Rahul Pathak, "Pan-Islamic Fundamentalism: Exporting Terror," India Today, Delhi, May 15, 1994, pp. 24-37; followed by Gupta, "Pakistan: Forging an Arab Connection," pp. 38-39, and Gupta, "Afghanistan: Getting Stung by the Stingers," pp. 40-45. Extensive details, to include names and groups, are in Sati Sahni, Kashmir Underground, Delhi: Har-Anand, 1999; less detailed but useful, Arjun Ray, Kashmir Diary: Psychology of Militancy, Delhi: Manas, 2003.

14. For background, Bidanda M. Chengappa, "Pakistan’s Secret Power: The Inter Services Intelligence," Indian Defence Review, vol. 15 no. 1, January-March 2000.

15. Statistics published by Indian Army; http://www.armyinkashmir.org/militant.html. It could logically be argued that the security forces have an interest in inflating the figures, but coroner’s inquest procedures followed to dispose of bodies do not support such a contention. Consequent to contacts, bodies are turned over to police, who coordinate with local authorities to identify deceased and arrange burial. Fieldnotes, August-September 2003.

16. Census of India, 1991. By the 2001 Census, this number had risen to over 138 million. http://news.bbc.co.uk/2/hi/south_asia/3634546.stm.

17. A series of blasts on March 12, 1993 killed over 250 and wounded more than 700; on August 25, 2003, a similar episode killed at least 52 and wounded over 150. For a detailed recounting of such actions through 1999, see S.K. Ghosh, Pakistan’s ISI: Network of Terror in India, Delhi: A.P.H. Publishing, 2000. For the second episode in Bombay, see the series of articles and sidebars contained in "Return of Terror," India Today, September 8, 2003, pp. 14-25; Sheela Raval, "The Iceberg’s Tip," India Today, September 15, 2003, pp. 17-18; Farheen Hanif, "New Faces of Terror," India Today, September 22, 2003, pp. 12-18; and Joanna Slater, "The Terror That Stalks India," Far Eastern Economic Review (FEER), Hong Kong, September 4, 2003, pp. 14-16. For further discussion of external influences, see M.J. Akbar, "The Roots of the Mumbai Blasts," FEER, September 4, 2003, p. 26.
18. See e.g. Geoffrey Mohan, "Pakistan Calls for International Mediation in Kashmir Conflict," The Los Angeles Times, February 6, 2002, p. A3.

19. Excellent discussion may be found in Praveen Swami, "The Kargil War: Preliminary Explorations," in K.P.S. Gill and Ajai Sahni, eds., Terror and Containment: Perspectives on India’s Internal Security, Delhi: Gyan, 2001, pp. 99-139. See also, Rajeev Sharma, Pak Proxy War: A Story of ISI, Bin Laden and Kargil, Delhi: Kaveri, 1999.

20. A body of literature is beginning to appear on the conflict. See e.g. Y. Bahl, Kargil Blunder: Pakistan’s Plight, India’s Victory, Delhi: Manas, 2000.
21. An effort to place this move in context is Maroof Raza, ed., Generals and Governments in India and Pakistan, Delhi: Har-Anand, 2001.

22. See Karan R. Sawhny, ed., Kashmir: How Far Can Vajpayee and Musharraf Go? Delhi: Peace Publications, 2001.

23. For both the latter attack and context, see K. Bhushan and G. Katyal, Attack on Parliament: Challenges Before the Nation, Delhi: A.P.H. Publishing, 2002.
24. V.K. Sood and Pravin Sawhney, Operation Parakram: The War Unfinished, Delhi: Sage, 2003. It is perhaps significant that even prior to the December attack on Parliament, India was already openly discussing striking against militant training facilities in Pakistan. See the series of articles and sidebars under "Should India Attack?" India Today International, October 29, 2001, pp. 10-33. This had spawned discussion of nuclear war possibilities; See M.V. Ramana and A.H. Nayyar, "India, Pakistan and the Bomb," Scientific American, December 2001, pp. 72-83.

25. Sreedhar, ed., Pakistan After 9/11, Delhi: Manas, 2003.

26. For ties with Taliban, see esp. Seymour M. Hersh, "The Getaway: Questions Surround a Secret Pakistani Airlift," The New Yorker, January 28, 2002, pp. 36-40. For nuclear links, among myriad possibilities, see e.g. David E. Sanger, "Atomic Ties Link North Korea and Pakistan," International Herald Tribune, Neuilly Cedex, November 25, 2002, p. 1.

27. Though the ruling National Conference (NC) – which had claimed an absolute majority at 57 of 87 seats in the 1996 elections – maintained a narrow plurality in the 2002 contest at 28 seats, the PDP joined its 16 seats with the Congress Party’s 20 to form a government (assisted by votes from the 23 ‘Others’). Overall turnout for the four rounds was 44.0 per cent. For details, see Ramesh Vinayak, "Redeeming Revolt," India Today International, September 30, 2002, pp. 14-19; and Vinayak, "Big Fall for Big Change," India Today International, October 21, 2002, pp. 12-18. Evaluation is in Yudhishtar Kahol, ed., Kashmir: Return of Democracy, Delhi: Anmol, 2003.

28. A useful selection: Tim McGirk, "Has Pakistan Tamed Its Spies?", Time, New York, May 6, 2002, pp. 32-34 (contains excellent map at p. 34); Douglas Frantz, "The Rogue to Fear is the one Following Orders," The New York Times, February 13, 2002; Isabel Hilton, "Letter From Pakistan: The General in his Labyrinth," The New Yorker, August 12, 2002, pp. 42-55; Ahmed Rashid, "Pakistan on the Edge," The New York Review of Books, October 10, 2002, pp. 36-39; and Raj Chengappa, "Rocky Road to Peace," India Today International, May 12, 2003, pp. 20-25.

29. A useful discussion of popular attitudes in Kashmir itself is Craig Whitlock, "In Kashmir, Patriotism Goes Only so Far," The Washington Post National Weekly Edition, January 21-27, 2002, p. 17. See also Mamta Rajawat, ed., Kashmir: Shadow of Terrorism, Delhi: Anmol, 2003; and Hari Om, R.D. Sharma, Rekha Choudhury, Jagmohan Singh, and Ashutosh Kumar, eds., Politics of Autonomy in Jammu and Kashmir, Jammu: Vinod, 1999.

30. Excellent for illustrating this point is Alexander Evans, "The Kashmir Insurgency: As Bad as it Gets," Small Wars and Insurgencies, vol. 11 no. 1, Spring 2000, pp. 69-81. For a more recent, depressing assessment, see Praveen Swami, "J&K: Decapitated Nightingale," South Asia Intelligence Review, vol. 2 no. 52, July 12, 2004, South Asia Terrorism Portal, www.satp.org; Histrionic at times but also valuable is Arundhati Roy, "How Deep Shall we Dig? Repression and Violence in India," Z Magazine, vol. 17 no.7-8, July/August 2004, pp. 70-75.
31. See e.g. my "At the Frontlines of the GWOT: State Response to Insurgency in Jammu," Journal of Counterterrorism and Homeland Security International, vol. 10 no. 1, 2004, pp. 38-46.
32. I round this up to 10.5 million for statistical purposes below, since current estimates are in this neighborhood.
33. Most accessible source for data is Majid Husain, Systematic Geography of Jammu and Kashmir, Delhi: Rawat, 2000.

34. Predictably, there is also a strong movement in historically Buddhist Ladakh for enhanced local autonomy, with this being resisted by those concerned at the possible break-up of the traditional "Jammu & Kashmir" entity that is based upon the pre-independence princely State. Indeed, all maps on public sale go to the extent, not only displaying the present State as coinciding with its pre-independence boundaries, but also of not showing the Line of Control (LoC) that delimits cease-fire arrangements with Pakistan. That portion of the original State controlled by Islamabad is designated Pakistan occupied Kashmir (PoK) by New Delhi, or, from the Pakistani position, ‘Free Kashmir’ (Azad Kashmir). For a discussion of larger issues, such as demands for regional autonomy, see Hari Om, Rekha Choudhury, Jagmohan Singh, and Ashutosh Kumar, eds., Burning Issues in Jammu and Kashmir Politics, Jammu: Jay Kay, 1999.

35. Statistics vary as per source. I have used the highest available to make the statistical "worst case." For civilians and militants killed, see the Indian Army web site published from J&K: www.armyinkashmir.org. It is updated regularly, though frequently in tardy fashion. It does not publish security force deaths. For these, the official figure released in Delhi, 380, is used since it is higher than the 338 published by South Asian Terrorism Portal www.satp.org in its weekly South Asia Intelligence Review: Weekly Assessments and Briefings, vol. 2 no. 48, June 14, 2004.

36. (836 + 1,447 + 380 = 2,663) ¸ 10,500,000 = 25.4, when multiplied by 100,000. The figure could be raised by subtracting Ladakh’s population from the denominator, but statistics are only available in the aggregate for the entire State.

37. See www.safestreetsdc.com/subpages/murdercap.html, accessed September 26, 2003. Washington, DC and Detroit lead the list. Next come (above 14.0:100,000): Baltimore, 38.3; Memphis, 24.7; Chicago, 22.2; Philadelphia, 19.0; Columbus (Ohio), 18.1; Milwaukee, 18.0; Los Angeles, 17.5; and Dallas, 15.8. Of course, the actual methodology for compiling murder rates does not lump into the count criminals and law enforcement officials who fall in the campaign. In a U.S. scenario, those J&K insurgents killed would generally be matched by those American criminals incarcerated, with law enforcement casualties comparatively small. If only civilian casualties are considered – i.e. true "murder" victims – the 8.0 (7.96):100,000 figure is in the same neighborhood as New York (7.3). This is not idle calculation, since police sources in J&K, supported by statistics, state that criminal activity per se is a minor concern. (Field work, August-September 2003) Hence, unlike e.g. Colombia, where insurgent violence pales when compared to criminal violence, in the J&K case, insurgent violence is the problem.

38. Arundhati Roy makes this case poignantly: "Recently, a young Kashmiri friend was talking to me about life in Kashmir. Of the morass of political venality and opportunism, the callous brutality of the security forces, of the osmotic, inchoate edges of a society saturated in violence, where militants, police, intelligence officers, government servants, businesspeople, and even journalists encounter each other, and gradually, over time, become each other. He spoke of having to live with the endless killing, the mounting ‘disappearances,’ the whispering, the fear, the unresolved rumours, the insane disconnection between what is actually happening, what Kashmiris know is happening, and what the rest of us are told is happening in Kashmir. He said, ‘Kashmir used to be a business. Now it’s a mental asylum’." Note the discrepancy between the "endless killing" and the actual figures. See The Hindu, Chennai, April 25, 2004.

39. Particularly useful is Ashok Malik, Uday Mahurkar, and Sandeep Unnithan, "Terrorism’s New Strategy," India Today International, pp. 12-19; accompanied in the same number (pp. 20-22) by Shishir Gupta with Rajiv Deshpande, "No Soft Options."
40. There is little difference in the socio-economic-political particulars of those recruited in J&K and those one finds providing the foot soldiers in conflicts from Nepal to Colombia to Ulster. This is the "gang dynamic" so evident in U.S. inner cities and is a departure from what we may label the classic insurgency dynamic as illustrated by people’s wars in China, Vietnam, Thailand, the Philippines, and others. Given the nature of the conflict, it is particularly useful to compare its particulars with those of Ireland in the early part of the 20^th Century. See e.g. Peter Hart, The I.R.A. at War 1916-1923, New York: Oxford University Press, 2003, as well as his The I.R.A. and its Enemies: Violence and Community in Cork 1916-1923, New York: Oxford University Press, 1998; also Michael Farry, The Aftermath of Revolution: Sligo 1921-23, Dublin: University College Dublin Press, 2000.
41. Though theoretically quite distinct from a ‘military leads’ approach, as invariably practiced by the U.S. when it deploys in expeditionary mode, sorting out particulars can often be a difficult and painful process. See e.g. Stephen Deakin, "Security Policy and the Use of the Military – Military Aid to the Civil Power, Northern Ireland 1969," Small Wars and Insurgencies, vol. 4 no. 2, Autumn 1993, pp. 211-27.
42. POTA was repealed in December 2004, though its most significant operative clauses were simultaneously inserted in the Unlawful Activities (Prevention) Act, 1967.
43. For full texts and discussion, see Arnab Goswami, Combating Terrorism: The Legal Challenge, Delhi: Har-Anand, 2002.

44. Fieldnotes, August 2003.

45. Ibid.
46. Ibid. According to materials provided by local officials, Poonch has a total of 178 villages, of which 168 are inhabited.
47. Ibid.
48. Ibid.
49. Figures and details may be found in R.K. Jasbir Singh, ed., Indian Defence Yearbook, Delhi: Natraj, issued annually. Particularly useful is the 1999 volume.
50. Ibid, September 2003.
51. As per above, best comparison is with the extensive SWAT capabilities now contained within the US police system nationwide. For the legal parameters, internal regulations, and numerous particulars of the Indian situation, see Rekha Chaturvedi, Manual of Para Military Forces in India, Delhi: Universal Law Publishing, 2001.
52. This status has been the subject of continuous formal and informal bargaining since the union with India but remains intact. Among other things, it limits employment and land ownership to citizens of the State, with additional limits upon who can achieve that status.
53. See e.g. Praveen Swami, "Jammu & Kashmir: On the Edge in Kishtwar," Frontline, Chennai, August 29, 2003, pp. 17-20; also, Ajmer Alam Wani, "Doda Bears Brunt of Neglect, Governments’ Apathy," The Kashmir Times, Jammu, August 24, 2003, p. 7.
54. Swami, "Jammu & Kashmir: On the Edge in Kishtwar", p. 20.
55. (51 + 66 + 35 + 101) = 252 ¸ 690,474 (x 100) = 36.6
56. It is, in fact, slightly lower than the murder rate of Brazil’s capital, Brasilia, 38:100,000, as briefed by Schlumberger Security Consultants on March 7, 2003 in Bogota. It is useful, within Latin America, to ascend from this example, taking only representative and well known cities presented in the same briefing: Sao Paulo, 56:100,000; Caracas, 70:100,000; Rio de Janero, 95:100,000; Cali, 118:100,000; San Salvador, 150:100,000; and Medellin, 185:100,000.
57. A frequently cited work that perpetuates this misunderstanding as to the role of counterinsurgency is D. Michael Shafer, Deadly Paradigms: The Failure of U.S. Counterinsurgency Policy, Princeton, NJ: Princeton University Press, 1988.
58. Valuable for consideration is John Newsinger, "From Counter-Insurgency to Internal Security: Northern Ireland 1969-1992," Small Wars and Insurgencies, vol. 6 no.1, Spring 1995, pp. 88-111.

59. Useful comparison may be made with Valery Tishkov, Chechnya: Life in a War-Torn Society, Berkeley: University of California Press, 2004; to delve into colonial history, with Alan Warren, Waziristan: The Faqir of Ipi and the Indian Army, New York: Oxford University Press, 2000.

60. See Amy Waldman, "India and Pakistan: Good Fences Make Good Neighbors," New York Times, July 4, 2004.

61. See Praveen Swami, "J&K: Operation Sarp Vinash - The Army Strikes Hard," South Asia Intelligence Review, vol. 1 no. 46, June 2, 2003, South Asia Terrorism Portal, www.satp.org.

62. See Zahid Hussain and Surinder Singh Oberoi, "New Face of Terror," India Today International, February 26, 2001, pp. 26-31; and the three article series, in the December 8, 2003 number of the same magazine: Sayantan Chakravarty, "Exclusive: Confessions of Captured Fidayeen," pp. 10-12; Indrani Bagchi, "Beyond Control," pp. 14-17; and Raj Chengappa, "Between the Lines," pp. 18-19.

63. See e.g. Uday Mahurkar, "New Theatre: The Murder Probe Reveals Gujarati Youth are Being Trained in Pakistani Terror Factories," India Today International, May 12, 2003, p. 26; Mahurkar and Sheela Raval, "Innocence Betrayed," India Today International, July 5, 2004, pp. 36-40.

64. See Ahmed Rashid, "Pakistan: Danger Mounts for Embattled General," Far Eastern Economic Review, January 8, 2004, pp. 12-16; as well as Eliza Griswold, "In the Hiding Zone: Pakistan’s Lawless Tribal Borderland has Become a Virtual Jihadi Highway," The New Yorker, July 26, 2004, pp. 34-42.

65. An All Parties Hurriyat Conference delegation met the then Prime Minister, Atal Behari Vajpayee, in New Delhi on January 23, 2004.