By BarathVector Editorial — 2026-08-07

As of 7 August 2026, 21:00 IST. This is the second of two parts. The first examined the Makkah Joint Defence Agreement signed on 7 August and concluded that the clause is thin, has already failed its only live test, and will not be honoured as written — and that this does not make it safe.

The argument that closed part one was that a pact does not have to be honoured to do damage. It only has to make a short war look more plausible to the people who would have to choose one.

This part is about those people: the condition of the army that would make that choice, the frame it has adopted to justify it, the story it has told about the last war and then come to believe, and the specific sequence by which an attack on Indian soil now escalates faster than it did in May 2025.

One conclusion should be stated at the outset, because it is the opposite of the fatalism that dominates the Indian conversation. The war is not inevitable, and nothing in this analysis says it is. What is true is narrower and worse.

The state of the army that would start it

Pakistan's military in August 2026 is fighting on five fronts.

It has been at war with Afghanistan since February, under an operation named Ghazab lil-Haq — Righteous Fury — with two corps committed and reports of Pakistani forces holding territory near Kandahar. The United Nations Assistance Mission in Afghanistan recorded at least 499 Afghan civilians killed and 1,216 injured between 1 October 2025 and 30 June 2026 in Pakistani air strikes, artillery fire and border clashes.

The Tehrik-i-Taliban Pakistan is running a nationwide campaign against the state. Separately, and it is a different phenomenon, sectarian jihadism is killing Pakistani civilians: the suicide bombing of the Shia Khadija Tul Kubra Mosque at Tarlai Kalan near Islamabad on 6 February 2026, which killed 31 worshippers and injured more than 169, was claimed by Islamic State – Pakistan Province, not by the TTP. Balochistan is in insurgency. Pakistan-occupied Kashmir has been in open revolt since June, with the Joint Awami Action Committee banned under anti-terror law, the internet shut down, mass arrests, and dozens killed ahead of the July assembly vote. And roughly 13,000 personnel are garrisoning Saudi Arabia in the middle of a live war.

Pakistan now sits first in the Global Terrorism Index 2026 — the first time it has held that position — with 1,139 deaths and its worst year since 2013. Sushant Sareen's field picture for the Observer Research Foundation is worse than the aggregate suggests: insurgents controlling highways, occupying towns and running checkpoints; security forces withdrawn to bunkers; and, on his account, ministers paying protection money to militants. The hybrid regime, he writes, has "only one tool in its arsenal — brute force."

The conventional inference is that an army in that condition cannot afford another war. Sareen's case is the opposite: that an army in that condition needs one. It has a counter-case, which part one supplied — when this same general staff had a chance to convert a treaty into a war in March, it chose six weeks of mediation instead. An army that reaches for the diversionary option is not the same army that declined one five months ago. The question is what would change it.

The relief Pakistan will not take

Set against that condition is a fact the Pakistani general staff never has to explain at home, because nobody asks.

India's demand has not moved in twenty years, and it is narrow. It is not territory. It is not the constitutional status of Pakistan-occupied Kashmir, whatever is said at the United Nations. It is not a change of government in Islamabad. It is that the cross-border terrorist infrastructure — the camps, the launchpads, the fundraising fronts, the men who run them — be dismantled. A Pakistani government that met that single demand would not be conceding a war aim. It would be closing down an apparatus that its own citizens are now dying to sustain.

And the return on closing it would be the largest economic relief available to Pakistan from any source. Trade across the Attari–Wagah line is near zero. Pakistan currently buys food, medicine and industrial inputs through third countries at a premium, freighted the long way round, when the cheapest and closest supplier of nearly all of it sits across a land border. No IMF tranche, no Gulf deposit and no Chinese rollover delivers anything comparable, because every one of those is a loan against future austerity and normalised trade is a permanent reduction in the cost of living. Pakistan spends $11.9 billion a year on defence while its solvency depends on a $7 billion Fund facility and $12 billion of other people's money parked in its central bank. The arithmetic is not difficult and it is not disputed by Pakistani economists.

That is the choice, and it is a choice. The apparatus is retained, and it is now consuming its host faster than it damages its target: 1,139 dead inside Pakistan in a single year, first place on the Global Terrorism Index, whole districts of Balochistan run by people the state cannot reach. Money that could have gone to a power grid or a school system goes to the men holding the highways and to the infrastructure that keeps India permanently hostile. The cancer is metastasising in the body that cultivated it.

None of which reads as irrational once the question is asked correctly. It is not the country that is choosing. It is the army — and for the army, the confrontation is not a cost. It is the justification for a budget, a political veto, a business empire and a claim on the state that no ordinary military in a solvent democracy could hold. Normalisation with India is the one outcome that would make the Pakistan Army answerable to Pakistan. That is precisely why it is the outcome the institution cannot allow, and it is the strongest version of Sareen's argument: an army whose position depends on the confrontation continuing will not end the confrontation, and may prefer to escalate it.

The religious frame, and the reason it is failing

The Pakistan Army has been explicit about the instrument it has chosen.

Field Marshal Asim Munir revived the two-nation theory in a speech to the Overseas Pakistanis Convention in Islamabad on 16 and 17 April 2025, insisting that Muslims are "distinct from Hindus in all aspects of life — religion, customs, traditions, thinking and aspirations." The Pahalgam attack followed on 22 April 2025. Indian commentary has repeatedly placed those two dates side by side. No connection between the speech and the attack has been established, and none is alleged here.

The naming is deliberate and consistent. Pakistan called the May 2025 conflict with India Marka-e-Haq, the Battle of Truth. It called the Afghanistan operation Ghazab lil-Haq. At the anniversary event at General Headquarters in May 2026, Munir described the May 2025 conflict as "a battle between two ideologies." Religious legitimation is now running on the eastern and western fronts at once, in the same vocabulary of haq.

The furthest instance came at a press conference in Rawalpindi on 31 July 2026. Asked about warming ties between India and Afghanistan, the Director General of Inter-Services Public Relations, Lieutenant General Ahmed Sharif Chaudhry, used the word kafir. He applied it to India and to the BJP-RSS, characterised the Afghan Taliban authorities as having aligned themselves with that camp, and closed by citing scripture forbidding believers from taking disbelievers as allies — a passage whose second clause carries the implication that a Muslim who does so has placed himself outside the community.

The word is the story, and reporting it in softer English would misreport what was said. Kafir translates as unbeliever. In South Asian political speech it is not a neutral theological category: it is a term of contempt, and it is the word that has gone before a great deal of the violence this subcontinent has seen. A serving three-star officer speaking as the army's official voice does not reach for it by accident, and the intent is not obscure.

That the exclusion of the Afghan Taliban arrives by implication rather than by direct labelling is not a mitigation. It is a construction. It is deniable in translation and unmistakable to the audience it was written for, which is the point of building it that way.

The senior public voice of the Pakistan Army used the vocabulary of religious exclusion from an official podium, about a neighbouring Islamic government, in the context of that government's relations with India — and it did so as an instrument of state policy, not as a private religious opinion.

The purpose, on Sareen's reading and on the plain logic of the situation, is that an external religious enemy is among the very few frames under which a militant insurgency, a Baloch separatism and a Kashmiri revolt might conceivably be subordinated to the army that is fighting all three.

It is not working, and that is the most dangerous finding in this analysis.

The TTP and several allied groups are Deobandi-aligned in doctrine. Deobandism itself is a mainstream Sunni school of jurisprudence headquartered at Darul Uloom Deoband in Uttar Pradesh, with many millions of adherents in India and elsewhere who have no connection whatever to militancy. What follows concerns Deobandi-aligned militant organisations, and nothing beyond them.

Religious exclusion aimed at the Afghan Taliban does not rally those organisations behind the Pakistan Army. It confirms the TTP's founding argument — that the Pakistani state is itself apostate — from the state's own mouth. The factions have indeed united, and they have united against Pakistan. Field reporting carried by Indian and regional outlets in 2025 described the TTP providing operational training to Baloch Liberation Army militants, including coordinated attack timing and tactics designed to overstretch security forces; that reporting has not been independently corroborated and is presented here as claim rather than established fact. What is not in question is the output. Between 30 January and 5 February 2026 the BLA ran Operation Herof II — roughly 40 attacks across 12 districts of Balochistan, including a prison break at Mastung that freed about 30 inmates.

And Baloch separatism is ethno-nationalist and largely secular. No quantity of anti-kafir rhetoric addresses a grievance about resource extraction and enforced disappearance.

A mobilisation strategy that is failing is more dangerous than one that is working, because the operator has to keep raising the dose. If two years of two-nation rhetoric, a religiously branded western war and the vocabulary of exclusion from the ISPR podium have not consolidated the factions, then the one instrument that might — an actual confrontation with India — becomes harder to rule out.

The narrative machine, and why it is now a planning input

What happened in the air over four days in May 2025 remains internationally contested. The fortnight that followed is not, and it is the part that matters here: Pakistan won the information contest decisively, and India lost it.

The campaign is documented rather than alleged. Dr Sameer Patil, writing for RUSI, describes "a tsunami of disinformation from Pakistani social media handles intended to sow confusion" — the Pahalgam attack framed as a false flag, a fabricated claim that Pakistani hackers had disabled 70 per cent of India's power grid, a deepfake of India's External Affairs Minister, and Chinese state media amplification throughout. Patil's verdict on India's response is unsparing: New Delhi was "often caught flat-footed," relying on individual fact-checkers rather than institutional machinery, while Indian television undermined its own government by broadcasting false claims about India's own campaign.

None of that would matter much if it had stayed abroad. It did not. Sareen's finding is the one that connects the propaganda to the war risk: "Having internalised the narrative that Pakistan worsted India during Operation Sindoor, Pakistanis have convinced themselves that they could actually prevail."

An information operation whose own authors come to believe it stops being propaganda and becomes a faulty intelligence estimate. That is the route by which shameless bragging turns into a planning assumption, and a planning assumption into a decision. It is also the mechanism that makes the Makkah pact matter: a general staff that has already persuaded itself it won the last round is exactly the audience for a piece of paper suggesting the next round would be internationally underwritten.

Borrowed relevance, and the date it expires

The second assumption in Rawalpindi's list is more fragile than the first, because it depends on one man.

Pakistan has spent fifteen months converting a single American intervention into a standing entitlement. It nominated Donald Trump for the Nobel Peace Prize in June 2025 for the May ceasefire, praising his "stellar statesmanship"; it announced a second nomination in October 2025 over Gaza; and Shehbaz Sharif has called him the "saviour of South Asia." India's position throughout has been that there was no mediation to award — that the ceasefire was arranged directly between the two militaries. Whatever the truth of that dispute, its consequence is not in doubt. Pakistan has built a diplomatic identity on an American intervention the other party says did not happen.

That identity has bought real things: a mediator's role in the Iran war, a hearing in Washington that survived the Afghan campaign, and the standing that put a Pakistani prime minister between the Saudi crown prince and the Turkish president in Makkah. It is also entirely borrowed. It rests on the personal disposition of one American president toward one Pakistani field marshal, and it is denominated in a currency with a known expiry date — the same administration is fighting a war with Iran in which Pakistan is useful, and utility of that kind ends when the war does.

The planning risk is not that Rawalpindi is wrong to think Washington intervened last time. It is that a wasting asset is being entered into the ledger as a permanent one, and any calculation that assumes an American rescue in 2027 or 2028 is assuming a man, a mood and a war that may none of them be there.

How it spirals

Assume the trigger is what it has been in every previous cycle: an attack, attributed to Pakistan-based groups, on Indian soil.

The machinery is being rebuilt for that, and here the sourcing has to be stated plainly and kept attached. Indian intelligence assessments, relayed through Indian outlets and without independent corroboration, report that the Pakistani establishment provided PKR 4 crore to Lashkar-e-Taiba to rebuild its Muridke headquarters — a claim first reported on 14 September 2025, with a stated completion deadline of 5 February 2026 that has passed without follow-up reporting. The same body of Indian intelligence reporting, again uncorroborated, describes a Lashkar leadership meeting in Pakistan-occupied Kashmir resolving to increase launchpads, position more than a thousand fighters at them, and raise infiltration attempts to at least twenty a day — a figure framed conditionally in the original reporting, against an assumed current rate of about five.

Independent of those figures, ORF's Kanchan Lakshman documents the structural picture: dispersal rather than destruction, a new Lashkar training facility called Markaz Jihad-e-Aqsa in Lower Dir, Jaish-e-Mohammed broken into smaller sites and fundraising through digital payment platforms under cover of flood relief, and a new Jaish women's wing. His conclusion is that "achieving kinetic success does not automatically translate into strategic containment, and Pakistan's jihadist architecture continues to regenerate."

From the attack, the ladder is short, and India has shortened it.

India has already spent its sub-military rungs. The Indus Waters Treaty is in abeyance, ties are downgraded, Attari–Wagah is closed and trade is near zero. In 2025 those steps bought two weeks of visible action before force was used. Next time there is nothing between the attack and the missile except water — and the only water escalation left is physical diversion rather than treaty suspension.

That rung deserves more scrutiny than Indian coverage gives it. India tendered the ₹2,600 crore Chenab–Beas Link Tunnel on 20 May 2026 and cancelled the tender three days later, on 23 May; it is accelerating Ratle, Pakal Dul, Kiru and Kwar. In May 2026 an arbitral award backed Pakistan's position on pondage limits, which India rejected on the ground that the court was illegally constituted. Pakistan has said that disruption of its water would be an act of war — and that is not merely a rhetorical position to be reported and passed over. It is a claim with a treaty text, an arbitral finding and 240 million people behind it, and it is the one Indian escalation rung where the adversary's threatened response is both stated in advance and, on the record so far, legally supported.

The next Indian response probably opens where the last one ended. Operation Sindoor struck nine camps with fewer than fifty munitions and no state targets on the first night. The doctrine announced afterwards commits India to treating any cross-border terror attack as an act of war, to disregarding nuclear blackmail, and to not differentiating between terrorists and the government that harbours them. Taken at face value, that last clause means state targets from the outset. Army Chief General Upendra Dwivedi said around 30 May 2026 that the present situation is "only a temporary cessation of hostilities" and that the forces are ready for Operation Sindoor 2.0.

Munir would find it hard to absorb a strike silently. His field marshal's baton, the Twenty-Seventh Amendment, and the Chief of Defence Forces post created by that amendment — assented on 13 November 2025, with his own appointment notified on 5 December — rest on a domestic story in which he answered India in May 2025. Pakistani doctrine reinforces the personality: quid-pro-quo-plus requires a response exceeding India's level. The 2025 template was some six hundred drones in a single volley across India's western front, most destroyed by a layered network of Akash, Spyder, MR-SAM and S-400 systems. Carnegie assesses that Pakistan will race to close the gap by sourcing more advanced Chinese weapons — J-35s, HQ-19, KJ-500.

And the pact will not stop any of it, which is the calculation Rawalpindi is most likely to get wrong. If the Makkah clause is meant to deter India — to make Delhi hesitate because a strike on Pakistan now notionally implicates a NATO member and the custodian of the two holy mosques — there is no evidence in Indian conduct or Indian doctrine that it will work. The doctrine announced after May 2025 commits India to disregarding nuclear blackmail, which is a far heavier deterrent than an unpublished clause three governments have already declined to invoke once. India's own response to the signing was four lines from a spokesman. The considered Indian position is not that the guarantors are frightening; it is that they are irrelevant to the question of whether India answers an attack on its own citizens. On the record of the past fifteen months, the next Indian response to a mass-casualty attack will be heavier than Operation Sindoor, not lighter, and it will not be sized against the guest list at Al-Safa Palace.

And then the branch point that nobody decides. If a Pakistani salvo kills Indian civilians in numbers — by intent, by interception failure, or by guidance error in a six-hundred-drone saturation attack — every restraint on the Indian side disappears and there is no political space for a ceasefire on day four. In 2025 Pakistan struck military targets, which let India argue the escalation burden had shifted while keeping its own response calibrated. A single miss removes that argument.

Beyond that lies the zone where May 2025 stopped: sustained air campaigns rather than raids, naval interdiction, ground movement. And above it sits a nuclear command structure that has changed in ways India has barely discussed. The Twenty-Seventh Amendment replaced the National Command Authority with a National Strategic Command under reduced civilian oversight, abolished the Chairman of the Joint Chiefs of Staff Committee, and left the commander recommended by the army chief. ORF's Rahul Rawat and Shivam Shekhawat conclude that "'nuclear restraint' is therefore more likely to be compromised during a crisis," and note that with tactical weapons under predominantly army control, battlefield commanders may gain autonomy. Pakistan has never defined the when, where and how of first use. Under compressed timelines that ambiguity does not strengthen deterrence; it means nobody knows where the line is until someone crosses it.

Where the Makkah pact actually enters

It enters twice, in opposite directions, and part one's conclusion holds in both.

It enters at Pakistan's counter-response, because Turkish and Saudi diplomatic cover is worth more in the first seventy-two hours than any clause. Ankara supplied exactly that in May 2025 without a treaty, and a treaty makes it easier to supply again and harder to withhold.

And it enters at the off-ramp, in India's favour. Riyadh and Ankara now have formal standing and a direct financial stake in restraining Islamabad. Türkiye has a defence-export relationship to protect. China has CPEC and the overwhelming majority of Pakistan's arms market. The UAE has just concluded a strategic defence framework with India. There are more capitals with money at risk in a South Asian war than at any point in two decades, and every one of them would want it stopped by day five. Three guarantors who do not wish to be called are three guarantors with a powerful incentive to end a crisis before anyone calls them.

That is the dual effect stated in part one, and it is why the pact is neither the encirclement Indian television described nor the irrelevance the sceptics describe. It probably raises the chance a war starts and lowers the chance it runs long.

What India should do about it

The Ministry of External Affairs has the tone right: acknowledge, decline to escalate, remind the wealthiest signatory what it stands to lose. The substantive work is procurement, and it is running — roughly ₹52,000 crore cleared by the Defence Acquisition Council on 3 July 2026 for anti-drone systems, electronic warfare, medium-range surface-to-air missiles and a jet-based loitering munition, a shopping list written directly from the lessons of May 2025.

Three things should be added.

Fix the information front. Patil's finding that India was caught flat-footed, dependent on individual fact-checkers while its own television channels broadcast falsehoods about its own campaign, is a documented institutional failure and it is cheap to correct. Pakistan converted a contested military outcome into a domestic victory in a fortnight, and then made the decision to believe it. The next round of that contest starts before the first shot.

Keep every Gulf channel open, separately, and make the restraining interest expensive to abandon. The UAE framework, the Saudi Strategic Partnership Council, and the quiet Turkish rapprochement — foreign office consultations resumed this year after four years in abeyance — are each worth more than any declaratory response to the pact. The guarantors' desire to see a crisis ended by day five is an Indian asset. It is worth cultivating in peacetime, because it cannot be created in week one.

And go into the doctrine with both eyes open. The hardening was earned. For three decades India absorbed attacks and answered with dossiers, démarches and restraint that was read across the border as weakness rather than as maturity; the record of that period is that patience bought nothing and cost lives. Choosing to treat a mass-casualty attack as an act of war is a defensible response to that record, not the recklessness it is routinely called. It is also not free. A declared automatic response deters right up to the moment it fails, and then it binds. The decision to fight has been pre-delegated — to whichever group achieves an operational success large enough to cross the line, at a time of its choosing rather than India's. India should hold the doctrine, and should build the capacity to choose the timing back: hardened attribution, pre-cleared target sets, and the political preparation to answer at a moment of its own selection rather than on the attacker's schedule.

The thing to watch

Not the photographs from Al-Safa Palace. Not the clause, which will not be honoured as written, and not the guarantors, who do not want to be called.

Over the next eighteen months, watch the infiltration rate along the Line of Control against the twenty-a-day figure Indian intelligence attributes to Lashkar. Watch the construction at Muridke, whose reported completion deadline passed in February without follow-up, and at Lower Dir. Watch whether Pakistan draws down from its Afghan front or from Saudi Arabia, because either would free capacity eastward. Watch the Chenab tunnel and the pondage dispute, because water is the one casus belli Pakistan has already named and the one where an arbitral tribunal has already found against India. And watch whether the religious frame keeps escalating, because the surest sign that the unification strategy has failed will be the army reaching for a stronger dose.

The war is not inevitable. What is true is narrower and worse: over the past fifteen months both governments have reduced their own ability to not escalate. They have not done so in the same way and the two should not be equated. India spent its sub-military options and committed publicly to automatic retaliation — a deliberate policy adopted in daylight, debated, and reversible by the government that adopted it. Pakistan abolished its civilian brakes, folded the nuclear command into a structure recommended by the army chief, and handed one man the army, the joint services and the strategic command. One is a country hardening its policy. The other is an institution removing the last people who could tell it no.

What the two have in common is only this: neither has much room left to step back, and the event that would start the fire is held by neither of them.