Border Disputes in South Asia

Almost every land boundary in South Asia was drawn by someone who was not going to live on either side of it. The region inherited its lines from a departing imperial power that had drawn them for administrative convenience, military advantage or cartographic tidiness, and it inherited them at a moment when the successor states were least able to survey, negotiate or compromise. The result is a region where the boundary itself — not merely what lies on either side of it — remains an open legal question in every direction except one.

Frontier and Boundary: The Distinction That Explains South Asia

• A frontier is a zone; a boundary is a line. The separation was given its classic statement by Lord Curzon in his Romanes Lecture on Frontiers, and it remains the sharpest tool for reading Himalayan disputes.
  • A frontier is a belt of territory of indeterminate width in which two political systems fade into each other — thinly governed, often mountainous, with populations owing shifting or dual allegiance.
  • A boundary is a precise line with a legal definition and a determinate location on the ground, on either side of which sovereignty is exclusive and complete.
  • Curzon’s own argument was that the transition from frontier to boundary is the signature of the modern state, and that the process is inherently violent because it converts ambiguity into a zero-sum claim.

“Frontiers are indeed the razor’s edge on which hang suspended the modern issues of war or peace, of life or death to nations.” — Lord Curzon

• The Himalayan boundary is a frontier that later states tried to read as a line. For most of its history the Tibet–India tract was a grazing and trading frontier administered through custom, monastic jurisdiction and revenue practice rather than through survey.
  • Successor states arriving in 1947 and 1949 required exclusive linear sovereignty, because that is the only thing the modern international system recognises.
  • Each then went into the historical record and found evidence supporting a line — and both genuinely found some, because a frontier generates documents pointing in several directions at once.
  • Hence talks have never been able to proceed by simply reading the treaty: there is no single treaty covering the whole alignment, only a patchwork of conventions, maps, revenue records and administrative practice.

Allocation, Delimitation, Demarcation, Administration

• Boundary-making runs through four stages, a sequence set out most influentially by Stephen B. Jones, and South Asia’s failures cluster in the last two rather than the first.
  • Allocation — the political decision that a territory belongs to state A rather than state B. Partition and the princely accessions largely completed this.
  • Delimitation — the description of the line in words and on a map. The Sugauli Treaty, the Radcliffe Award and the Simla Convention are delimitation instruments.
  • Demarcation — the physical marking of the line on the ground with pillars or cairns, by joint survey.
  • Administration — continuing maintenance, joint inspection, and the running of the line as a working institution.
• The value of the sequence is that it locates the fault precisely.
  • Sir Creek is a delimitation dispute about the meaning of a 1914 resolution, and therefore admits a legal answer.
  • Kalapani is a delimitation dispute compounded by a demarcation failure — the treaty names a river, the river’s source was never surveyed, and no pillars were ever placed.
  • Siachen is a pure demarcation gap: the line was delimited to a grid reference and then simply stopped.
  • The Line of Actual Control fails at the level of administration — there is no agreed alignment to maintain, so patrol practice itself becomes the dispute.

Stage	What it produces	Examples	Where it broke
Allocation	Political award of territory	Partition 1947; princely accessions	Largely completed; Kashmir the exception
Delimitation	Treaty text and map	Sugauli 1816, Radcliffe 1947, Simla 1914	Ambiguous drafting; unratified instruments
Demarcation	Pillars on the ground	India–Bangladesh; India–Myanmar	Never done at Siachen, Kalapani, most of the LAC
Administration	Joint inspection and maintenance	Boundary working groups	No agreed line to administer on the LAC

South Asia’s boundary disputes are overwhelmingly failures of demarcation and administration, not of allocation — which is why they are technically soluble and politically frozen.

The Colonial Manufacture of a Disputable Border

• Colonial boundary-drawing followed strategic, not social, logic. The British Indian empire drew outward lines to manage buffers and inward lines to manage revenue, and neither purpose required the line to survive a court a century later.
  • The Durand Line of 1893 was drawn to fix the limit of British responsibility toward Kabul, not to divide a nation — it runs straight through Pashtun tribal territory.
  • The McMahon Line of 1914 was drawn to secure a strategic watershed against Chinese expansion into Tibet, at a conference whose Chinese plenipotentiary initialled but whose government repudiated.
  • The Radcliffe Award of 1947 was completed in about five weeks by a man who had never visited India, working from outdated maps with no time for survey.
• Unilateralism is the recurring structural feature. Several of the most contested alignments were fixed by one party acting alone, and the other side’s successor state has never accepted the procedure.
  • The 1914 Bombay Government Resolution on Sir Creek was an act of a provincial government resolving a dispute between a British province and a princely state, not an international treaty.
  • The Johnson Line in Aksai Chin was drawn by a survey officer on his own initiative and was never communicated to, still less accepted by, China.
  • Nepal’s western boundary was fixed by a treaty imposed after defeat in war — one reason Kathmandu treats revision as a matter of national dignity.
• Inadequate survey is the central defect, not an incidental one. Nineteenth-century triangulation stopped where the terrain became lethal, so the most disputed stretches are precisely the ones never fixed.
  • The source of the Kali river was never determined by survey — the entire Nepal dispute lives in that gap.
  • The alignment north of NJ9842 was never traced, because the 1949 negotiators assumed no army would ever go there.
  • Aksai Chin was mapped from a distance if at all, and India’s own official maps of it were revised as late as 1954 — a point A.G. Noorani presses hard.
• Rivers were used as boundaries in a landscape where rivers move. The Gangetic and Brahmaputra plains carry rivers that migrate laterally by kilometres in a decade, so a boundary defined as “the river” is a boundary that walks.
  • Susta on the Gandak is the classic instance; the riverine sectors of the Bengal border produced the same problem at larger scale, with char lands emerging and vanishing between surveys.
  • The legal fix — thalweg for navigable rivers, median line for others, and a fixed line where the channel jumps its bed — was available but was rarely written into the original instruments.
• Population was moved without regard to the line and the line without regard to population. The Cooch Behar enclaves are the extreme case — a chessboard of sovereign fragments produced by pre-colonial treaties, frozen by Partition into an absurdity that took sixty-eight years to undo.

The Legal Architecture: What Governs a Boundary

Inheritance — Uti Possidetis Juris

• New states inherit the predecessor’s lines. The principle of uti possidetis juris — “as you possess under law” — converts internal colonial boundaries into international ones at the moment of independence.
  • It was developed in Latin American decolonisation and generalised by the International Court of Justice in the Frontier Dispute case between Burkina Faso and Mali as a principle of general application.
  • Its purpose is stability, not justice: it freezes an arbitrary line precisely to prevent successor states from litigating every arbitrary line at once.
  • India relies on it comprehensively — Radcliffe, Sugauli and McMahon are all defended as inherited titles rather than as fair lines.
• The principle cuts both ways. Nepal invokes Sugauli against India and Pakistan invokes the 1914 Resolution against India; Afghanistan is the one state in the region that rejects the principle outright.

Possession — Effectivités, Acquiescence and the Problem of Maps

• Effectivités are acts of state authority on the ground — administration, revenue collection, policing, census enumeration — offered as evidence of title.
  • The Court’s settled position is that where a title is clear, title prevails over effectivités; where the title is ambiguous, effective administration decides.
  • This is India’s argument at Kalapani: the treaty does not resolve the source of the river, but Pithoragarh revenue records show Indian administration into the nineteenth century.
  • It is also China’s argument in Aksai Chin — whatever the maps said, the Xinjiang–Tibet highway was built through it and administered without effective Indian protest.
• Acquiescence and estoppel do a great deal of work in boundary law. A state that receives a map, understands its implications and says nothing may be held to have accepted it.
  • The classic authority is the Court’s judgment in Temple of Preah Vihear, where Thailand’s decades of silence about an annexed map defeated its later reliance on the watershed rule the treaty actually stated.
  • That reasoning is why cartographic protest is not a ritual: every objection India files against a Chinese renaming exercise or a Nepali map is an act of preserving the record against estoppel.

Rivers, Mountains and the Sea

• The thalweg rule governs navigable rivers. The boundary follows the line of the deepest continuous channel rather than the geometric middle, so that neither riparian is shut out of the usable channel.
  • Its corollary is that the boundary moves with slow accretion but not with sudden avulsion, when the river jumps its bed.
• The median line governs non-navigable rivers, running down the geometric centre between the banks.
• The choice between them is worth territory, which is why Sir Creek turns on whether the creek is navigable — the two rules place the line in materially different positions.
• The watershed principle places a mountain boundary along the crest dividing river systems, because it is visible, stable and hydrologically meaningful.
  • It is the stated basis of the McMahon Line, of the 1890 Anglo-Chinese Convention on the Sikkim–Tibet boundary, and of India’s argument that the line north of NJ9842 must follow the Saltoro Ridge.
  • Its weakness is that a complex range has more than one watershed, and each party selects the one that suits it.
• Customary boundaries rest on long grazing, pilgrimage and revenue practice, and are strongest where both parties’ local records corroborate them — as with Bhutanese pastoral use of Doklam.

The Sea — UNCLOS and the Turn to Equity

• The territorial sea is delimited by equidistance under the Law of the Sea Convention, unless historic title or special circumstances require otherwise.
• The EEZ and continental shelf are delimited so as to “achieve an equitable solution” — a standard, not a formula, and deliberately so.
• The method is a disciplined three-stage exercise, applied consistently by courts and tribunals.
  • Draw a provisional equidistance line from the relevant coasts.
  • Adjust it for relevant circumstances — most importantly coastal concavity, which cuts off a state sitting in the middle of a curve.
  • Apply a disproportionality check comparing each state’s share of the delimited area against its share of relevant coastline.
• This method settled the Bay of Bengal, and its availability is the single most important reason the sea was settled while the land was not.

Principle	What it decides	Where it operates in South Asia
Uti possidetis juris	Colonial lines become international	Radcliffe, Sugauli, McMahon; rejected by Afghanistan on the Durand Line
Effectivités	Administration as evidence of title	Kalapani revenue records; the Aksai Chin highway
Acquiescence	Silence as acceptance	Why India protests every renaming and every new map
Thalweg	Deepest channel, navigable rivers	Sir Creek — India’s case
Median line	Geometric centre, non-navigable rivers	Sir Creek — the fallback if navigability fails
Watershed	Crest dividing river systems	McMahon Line; Saltoro Ridge; the 1890 Convention at Doklam
Equidistance adjusted for equity	Maritime zones	The 2014 Bangladesh award; every settled Indian maritime boundary

Third-Party Settlement and India’s Guarded Posture

• South Asia has used adjudication sparingly on land and successfully at sea, and India’s formal legal position explains why.
  • India’s declaration accepting the compulsory jurisdiction of the International Court of Justice, deposited on 27 September 2019, carries reservations that between them remove every boundary question from the Court’s reach.
  • It excludes disputes with any present or former member of the Commonwealth, which covers Pakistan, Bangladesh and Sri Lanka.
  • It excludes disputes concerning “the modification or delimitation of its frontiers or any other matter concerning boundaries”.
  • It separately excludes disputes over “the determination and delimitation of its maritime boundaries“.
• The record is better than the reservations suggest. India has submitted to third-party process three times on boundary or resource questions and accepted the outcome each time — the Rann of Kutch tribunal of 1968, the Kishenganga arbitration, and the Bay of Bengal award of 2014, the last of which went substantially against it.
• The 2019 declaration is a decision about the future, not a description of the past. India accepted an adverse maritime award and then, five years later, formally closed the door to being taken to the Court on any boundary at all.

India’s Land Frontier: The Ledger

• India’s land boundary runs to about 15,107 km across seven neighbours, and its status varies more sharply than any comparable country’s.

Neighbour	Length	Status of the boundary	Indian states
Bangladesh	4,096.7 km	Settled — 1974 agreement, implemented 2015	West Bengal, Assam, Meghalaya, Tripura, Mizoram
China	3,488 km	Undelimited; no agreed LAC alignment	Ladakh, Himachal, Uttarakhand, Sikkim, Arunachal
Pakistan	3,323 km	Boundary settled except Sir Creek; LoC and AGPL are military lines	Gujarat, Rajasthan, Punjab, J&K, Ladakh
Nepal	1,751 km	About 98% agreed; Kalapani and Susta open; open border	Uttarakhand, UP, Bihar, West Bengal, Sikkim
Myanmar	1,643 km	Settled — 1967 agreement; being fenced	Arunachal, Nagaland, Manipur, Mizoram
Bhutan	699 km	Settled; open border	Sikkim, West Bengal, Assam, Arunachal
Afghanistan	106 km	Along Pakistan-occupied territory; not administered by India	Ladakh

• Three patterns are visible at once. The settled land boundaries are with Bangladesh and Myanmar; the open borders are with Nepal and Bhutan, a settlement so complete that no fence is thought necessary; and the unsettled ones are with the two states India has fought wars against.

The China Boundary as a Boundary Question

The India–China boundary is the only stretch in the region where the parties do not agree even on a line to disagree about. India has no boundary treaty with China anywhere along the alignment, which makes this a dispute at the delimitation stage that has been managed for six decades as though it were a policing problem.

The Western Sector: Johnson, Ardagh and Macartney–MacDonald

• Aksai Chin is a dispute between two British-era proposals that were never reconciled.
  • The Johnson Line, drawn by a survey officer in 1865 and later associated with Ardagh, ran along the Kunlun range and placed the whole of Aksai Chin inside Kashmir. It was never communicated to China and rested on no agreement.
  • The Macartney–MacDonald Line, offered to Beijing in 1899, ran along the Karakoram watershed and conceded most of Aksai Chin to China as the price of a settled frontier. China never replied.
  • British India therefore held two incompatible official positions on the same ground, and the successor states each adopted the one that favoured them.
• China’s effectivité is the highway. The Xinjiang–Tibet road was built across the plateau in the mid-1950s, and its discovery converted a cartographic disagreement into a live territorial dispute.
• The plateau’s value is that it is worthless as terrain. Nehru’s dismissal of it in Parliament remains the most quoted sentence in Indian boundary politics — and it was answered from the floor by a member who pointed at his own bald head and asked whether it should therefore be cut off.

“Not a blade of grass grows there.” — Jawaharlal Nehru, on Aksai Chin

The Eastern Sector: Simla, McMahon and Tawang

• The McMahon Line was delimited at Simla in 1914 and has never been accepted by China.
  • It runs roughly 890 km along the highest watershed, from the Bhutan trijunction to the Myanmar border.
  • It was drawn in bilateral notes between British India and Tibet, so its status depends on whether Tibet had treaty-making capacity — the question on which the entire eastern sector turns.
  • The Chinese plenipotentiary initialled but did not sign, and Beijing repudiated the Convention on the ground that Tibet was not a sovereign party.
• India’s answer is that Tibet exercised treaty capacity in fact, that China’s conduct for decades was consistent with the line, and that the alignment follows the natural watershed and the limit of effective administration.
• Tawang is the sharpest point. It carries a monastery of first-order significance in Tibetan Buddhism and is the birthplace of the sixth Dalai Lama, giving China’s claim a legitimacy dimension it has nowhere else.
  • India’s counter rests on the 2005 Agreement on Political Parameters and Guiding Principles, whose Article VII commits both sides to safeguard settled populations in the border areas — read by India as excluding populated Tawang from any package.

From Boundary to Line of Actual Control

• The LAC is not a boundary and was never meant to be one. It describes where each side’s forces actually are, and entered treaty language only in the 1993 Agreement on the Maintenance of Peace and Tranquillity.
• It runs through three sectors with different characters and different failure modes.
  • The western sector, about 1,597 km in Ladakh, carries the largest divergence between the two sides’ claims and holds Depsang, Galwan, Pangong Tso and Demchok.
  • The middle sector, about 545 km in Himachal Pradesh and Uttarakhand, is the least contested and the only sector in which maps were exchanged, in 2002.
  • The eastern sector, about 1,346 km along Sikkim and Arunachal, is where the McMahon question sits and China claims roughly a whole Indian state.
• The map-exchange failure of 2002–03 is the pivot of the process. When the parties moved to the western sector each refused to accept the other’s depiction, and the exchange has never resumed — leaving a management architecture with nothing to manage against.
• That architecture is real and has largely worked — the agreements of 1993 and 1996 on peace and military confidence-building, the 2005 protocol on modalities, the Working Mechanism for Consultation and Coordination of 2012, and the Border Defence Cooperation Agreement of 2013.

Cartography, Renaming and the Politics of the Map

• China has issued successive batches of Chinese names for places in Arunachal Pradesh, styling the state “Zangnan” or South Tibet, in repeated rounds with the largest batches in recent years.
• The exercise is legally instrumental, not merely symbolic — its purpose is to build a record of assertion which, if unanswered, could later be offered as evidence of an uncontested claim.
• India’s response is deliberately formulaic, because the point of a formula is that it is repeated identically every time.

“Assigning invented names does not alter the reality that Arunachal Pradesh was, is, and will always remain an integral and inalienable part of India.”

— Ministry of External Affairs

Galwan, the 2024 Understanding and the Thaw

• The Galwan clash of June 2020 killed twenty Indian soldiers and ended a working assumption — that the LAC could be managed indefinitely without being defined.
  • It produced the first combat fatalities on the boundary since 1975, and showed that protocols forbidding firearms did not prevent lethal violence.
  • Staged disengagement followed at Pangong Tso, Gogra and Hot Springs, each leaving buffer zones in which neither side patrols.
• The October 2024 patrolling understanding restored patrolling at Depsang and Demchok and unlocked a wider normalisation — but it is a patrolling arrangement, not a boundary settlement. It says where soldiers may walk, not where the line is.
• The 2025–26 thaw has been real but bounded. Special Representatives’ talks resumed, direct flights and the Kailash Mansarovar pilgrimage restarted, and the relationship warmed — none of it touching delimitation.

The Pakistan Boundary: Three Distinct Problems

• The India–Pakistan frontier is three legal objects with three different characters, and conflating them is the commonest error in analysing it.
  • The International Boundary from Gujarat to Jammu — settled by the Radcliffe Award, demarcated, and disputed only at its southern terminus in Sir Creek.
  • The Line of Control in Jammu and Kashmir — a military line with a treaty pedigree but no status as an international boundary.
  • The Actual Ground Position Line north of NJ9842 — a line of contact never delimited, demarcated or jointly authenticated at all.

The Line of Control: Karachi 1949 to Simla 1972

• The Karachi Agreement of 27 July 1949 created the ceasefire line after the first war, describing it in detail on the ground and placing it under United Nations military observers.
• The Simla Agreement of 2 July 1972 converted it into the Line of Control and, more importantly, changed its legal character.
  • Both sides undertook to respect the line resulting from the ceasefire of 17 December 1971 “without prejudice to the recognised position of either side“.
  • Both undertook to settle differences bilaterally, which India reads as excluding mediation and internationalisation.
  • Neither the line’s status nor the territory’s status was determined — Simla is a conflict-management instrument, not a boundary treaty.
• A line stable in fact and unsettled in law: heavily fenced and heavily manned, subject to a ceasefire understanding renewed in 2003 and reaffirmed in 2021, yet incapable of being cited as a boundary by either party.
• The events of 2025 did not change that character. The Pahalgam attack of 22 April 2025, Operation Sindoor on 7 May and the ceasefire reached through the directors-general of military operations on 10 May were an exchange across the line, not about it.

Where the Lines Meet: Gilgit-Baltistan, Shaksgam and the 1963 Agreement

• The northern end of the Pakistan boundary is also the start of the China boundary, and one instrument connects them.
  • By the Sino-Pakistan Boundary Agreement of 2 March 1963, Pakistan recognised Chinese sovereignty over the Shaksgam Valley — the Trans-Karakoram Tract of roughly 5,180 square kilometres north of the Siachen region.
  • India treats the cession as void: Pakistan cannot transfer territory it does not lawfully hold, and Parliament’s resolution of 22 February 1994 declares the whole of Jammu and Kashmir, including the areas under Pakistani and Chinese occupation, an integral part of India.
  • The agreement itself concedes the point. It provides that the parties will reopen negotiations with the sovereign authority of Kashmir once that dispute is settled — an express admission that Pakistan’s title is provisional.
• The cession is why Siachen matters strategically. It gave China a boundary terminus at the Karakoram Pass, precisely the point to which Pakistan wants the line from NJ9842 drawn — and India’s objection is that the Pakistani reading would produce a Sino-Pakistani land junction across the Saltoro.
• Gilgit-Baltistan’s constitutional limbo mirrors the LoC’s legal limbo. Administered by Pakistan since 1947 but never formally integrated, it was offered “provisional provincial status” rather than full provincehood — because full integration would concede the plebiscite argument Pakistan has built its Kashmir case on.

Sir Creek: The Region’s Sharpest Legal Set-Piece

Sir Creek is a 96-kilometre tidal channel in the marshes of the Rann of Kutch, once called Ban Ganga, dividing Gujarat from Sindh before it opens into the Arabian Sea. It is uninhabited and marginal in itself, and it is the most consequential unresolved boundary question India has, because it determines a maritime boundary worth thousands of square kilometres.

• Pakistan’s case rests on the 1914 Resolution settling a dispute between the Government of Sindh and the Rao Maharaj of Kutch.
  • Paragraphs 9 and 10 are read as placing the whole creek within Sindh, fixing the boundary along the eastern bank — the “ribbon” or Green Line. On that reading Pakistan owns the entire channel and Indian territory begins at the far shore.
• India’s case is that the boundary runs mid-channel, confirmed by later practice.
  • The Resolution’s accompanying map and the 1925 map show a mid-channel alignment, and mid-channel pillars were installed in 1924.
  • The thalweg principle applies because the creek is navigable at high tide, and under it the boundary must follow the deepest channel.
  • Pakistan’s rejoinder is precisely that the creek is not navigable — it is tidal marsh — and that the thalweg rule therefore has no application.
• The stakes are almost entirely offshore, because of what a terminus point does to a maritime boundary.
  • The land boundary’s endpoint is the origin of the baseline from which the territorial sea, the contiguous zone, the exclusive economic zone and the continental shelf are all measured.
  • A displacement of a few kilometres at the creek’s mouth fans outward across 200 nautical miles, translating into thousands of square kilometres of seabed and water column.
  • The area is among the region’s richest fishing grounds and is believed to hold hydrocarbon potential neither state can develop while title is contested.
• The negotiating history is complete and barren.
  • The Rann of Kutch tribunal of 1968, convened after the 1965 war on British persuasion, awarded Pakistan about ten per cent of a claim to some 9,000 square kilometres — but Sir Creek itself was left out of its remit.
  • Twelve rounds of talks since 1969 produced a joint hydrographic survey and nothing else; the last substantive round was in 2012.
• The absence of a maritime boundary has a security cost. Fishermen of both countries are detained after straying across a line neither can locate, and the 26/11 attackers reached Mumbai by sea through these waters.

Question	India’s position	Pakistan’s position
Governing instrument	1914 Resolution read with the 1925 map and 1924 pillars	1914 Resolution alone, paragraphs 9 and 10
Where the line runs	Mid-channel	Eastern bank — the “ribbon” or Green Line
Is the creek navigable?	Yes, at high tide — so thalweg applies	No — tidal marsh; thalweg inapplicable
Nature of the instrument	A provincial settlement, later refined by survey	A binding delimitation of the boundary
What is really at stake	The baseline, and therefore the EEZ	The same — hence the refusal to sever land from sea
Preferred sequence	Settle the maritime boundary first, or in parallel	Settle the creek first, on the 1914 reading

Siachen: NJ9842 and “Thence North to the Glaciers”

• The dispute is the product of four words in a 1949 document. The Karachi Agreement traced the ceasefire line in detail up to grid reference NJ9842 and then, because the negotiators assumed the terrain beyond was uninhabitable and militarily irrelevant, described the remainder as running “thence north to the glaciers“.
• Two readings are grammatically available.
  • Pakistan reads it as a straight line north-east to the Karakoram Pass on the boundary with China, which places the entire glacier system on its side.
  • India reads it as following the watershed north along the Saltoro Ridge — the mountain rule, and the alignment India in fact holds.
• India pre-empted the argument physically. Operation Meghdoot, launched in April 1984 after Pakistan began issuing permits for foreign expeditions to the glacier, put Indian troops on the Saltoro Ridge first — and possession of the ridge crest, not of the glacier, is what has held since.
• The strategic case has three limbs.
  • The Saltoro heights separate Pakistani forces in the Northern Areas from Chinese positions in the Shaksgam valley, denying the two a physical junction.
  • They give observation over Gilgit-Baltistan and the approaches used by the China–Pakistan Economic Corridor.
  • Dominating ground gives India the stronger bargaining position in any eventual settlement.
• The deadlock is a sequencing problem with a precise shape.
  • India’s position is that the Actual Ground Position Line must be delineated, authenticated and annexed to an agreement before any withdrawal, so that vacated heights cannot later be claimed or occupied.
  • Pakistan’s position is that authentication would legitimise a line created by force, so withdrawal must come first, to positions held before 1984.
  • The Indian Army’s institutional objection is memory: Kargil in 1999 demonstrated what happens when heights are vacated, and no service chief has since accepted withdrawal without a signed line.
• Siachen was one of the eight baskets of the Composite Dialogue, handled at defence-secretary level, and came closer to settlement in the mid-2000s than before or since.

“Fighting for Siachen is like two bald men fighting over a comb.” — Stephen P. Cohen

The Environmental Register: Glacier, Waste and the Peace Park

• Siachen is the highest militarised zone on earth, and its ecological cost is a distinct question from its strategic one.
  • The overwhelming majority of Indian fatalities on the glacier have come from altitude, cold, crevasse and avalanche rather than from enemy fire — a fact that reframes the deployment as an environmental problem as much as a military one.
  • The Gayari avalanche of 2012 buried a Pakistani battalion headquarters and killed some 140 people, prompting the most serious Pakistani call for demilitarisation in the dispute’s history.
• The glacier is retreating and thinning, in line with the wider Karakoram and Himalayan pattern; the deployment is an aggravating local factor rather than the primary cause.
  • Sustained military activity generates non-biodegradable waste, chemical and fuel residue and human waste that does not decompose at those temperatures, and which is often tipped into crevasses, entering the Nubra and Shyok drainage and eventually the Indus.
  • Both armies run waste-extraction programmes, which mitigate the symptom and leave the cause untouched.
• Environmental peace-keeping proposes that shared ecological stakes can carry cooperation where security bargaining cannot.
  • The scholarly frame is environmental peacemaking, associated with Ken Conca and Geoffrey Dabelko — ecological interdependence builds the confidence and working habits that later spill into harder issues.
  • The concrete proposal is a transboundary peace park of the kind surveyed by Saleem H. Ali — demilitarising the glacier and making it a jointly managed protected area with scientific access.
  • The idea has institutional history — proposals for a Siachen Science Centre and a high-altitude research station, and Manmohan Singh’s call, made at the base camp itself, for the glacier to become a “mountain of peace” rather than a symbol of hostility.
• Why it has not happened is not environmental.
  • A peace park requires demilitarisation, which returns to the authentication question — India will not vacate the Saltoro Ridge without a signed line, and Pakistan will not sign one.
  • Environmental cooperation is easiest where stakes are low and hardest where terrain is strategic; Siachen is the least favourable possible case for the theory, because the environmental asset and the strategic asset are the same object.
  • The realistic intermediate steps are correspondingly modest: joint glaciological monitoring, shared meteorological and avalanche data, coordinated waste extraction, and scientific access under existing military control — none of which requires either side to move a soldier.

Environmental peace-keeping at Siachen fails not because the ecological case is weak but because the ridge that must be vacated to build the park is the only thing either side is fighting for.

The Nepal Boundary: Which Stream Is the Kali?

• The Nepal dispute is about hydrography. The Treaty of Sugauli of 1816, concluded after the Anglo-Nepalese War, fixed Nepal’s western limit at the Kali river — also called the Mahakali or Sarada — and required Nepal to renounce everything west of it.
  • Both states accept the treaty. Both accept that the Kali is the boundary. Neither disputes the text.
  • What they dispute is which of several headwater streams is the Kali, because the treaty names a river whose source was never surveyed and never marked.
• Nepal’s case is that the Kali rises at Limpiyadhura, north-west of the Lipulekh pass. On that reading the whole triangle containing Kalapani, Lipulekh and Limpiyadhura — roughly 335 square kilometres — lies east of the river and therefore in Nepal, in Darchula district of Sudurpashchim province.
• India’s case is that the Kali rises at Kalapani, well below the pass, and that the stream from Limpiyadhura is a tributary rather than the main channel.
  • India adds an effectivités argument: whatever the hydrography, administrative and revenue records from the nineteenth century show the tract enumerated and taxed as part of Pithoragarh district.
  • Both governments produce British-era maps supporting their own reading — the expected result when the underlying survey was never done.
• The strategic value explains the intensity. Kalapani sits above 20,000 feet at the India–Nepal–China trijunction, has hosted Indo-Tibetan Border Police posts since 1962, and commands Lipulekh — a pass into Tibet and the route of the Kailash Mansarovar pilgrimage.

The 2020 Escalation and the Constitutionalisation of a Claim

• Two Indian acts triggered the crisis. India’s revised political map after the reorganisation of Jammu and Kashmir in November 2019 showed Kalapani inside Uttarakhand; and in May 2020 the Defence Minister inaugurated an 80-kilometre link road to Lipulekh through the disputed tract.
• Nepal responded by amending its Constitution. In June 2020 both houses voted unanimously to amend the schedule carrying the national emblem, embedding a new political map showing Kalapani, Lipulekh and Limpiyadhura as Nepali territory.
• The constitutional route makes this dispute structurally harder than any other on India’s borders.
  • A claim written into a constitution by unanimous vote can be retracted only by another amendment, requiring a two-thirds majority in a fragmented legislature.
  • It converts a negotiable technical question into a fixed national position, removing from every future government the ability to compromise even if it wished to.
  • The mirror image is instructive. India needed the 100th Constitutional Amendment to settle the Bangladesh boundary; Nepal used a constitutional amendment to make its own boundary claim unsettleable. The same instrument, used in opposite directions.
• A further irritant is the 2015 understanding to expand border trade through Lipulekh, agreed without Nepal, which Kathmandu protested as two states disposing of ground it claims.
• The turnover of 2025–26 has not changed the claim. The Gen Z uprising of September 2025, the interim government under Sushila Karki, the March 2026 election and Balendra Shah’s premiership altered everything in Nepali politics except this — the constitutional map stands, and no party has proposed touching it.

Susta and the Shifting River

• Susta on the Gandak is the second unresolved sector, and a textbook instance of the moving-river problem.
  • The boundary was defined by the Gandak (Narayani) river; the river has since migrated westward, leaving the settlement on the Indian bank while Nepal continues to claim it.
  • Roughly 145 square kilometres are in contention, and the legal question is whether the boundary is fixed at the river’s historic course or moves with the current channel.
  • The law has an answer — a boundary follows gradual accretion but not avulsion — but applying it requires an agreed historical channel, which no survey established.
• The Terai repeats the problem at lower intensity along its whole length — pillar loss, encroachment and “no man’s land” cultivation wherever a river forms the line.

What the Boundary Machinery Actually Achieved

• The technical work succeeded and remains unratified.
  • The Joint Technical Level Boundary Committee, working from 1981 to 2007, agreed roughly 98 per cent of the alignment and produced 182 strip maps, signed by the surveyors of both countries.
  • They have never been ratified by either government, because ratification would require settling the two sectors left out — Kalapani and Susta.
  • The Boundary Working Group of 2014 continues pillar construction and survey along the agreed stretches.
• A boundary can be technically finished and politically unfinished at once, and the residual two per cent holds the other ninety-eight hostage.

The Bhutan Boundary and India’s Security

The India–Bhutan boundary is settled and open. The Bhutan–China boundary is neither, and it is the only boundary negotiation in the world to which India is not a party but in whose outcome it has a direct and acknowledged security interest.

The Disputed Sectors

• China claims roughly 764 square kilometres of Bhutanese territory, and the strategic significance of the two main blocks is wholly asymmetric.

Sector	Area	Components	Why it matters
North-west	269 sq km	Doklam (89), Sinchulumpa (42), Shakhatoe (138)	Overlooks the Chumbi Valley and the approaches to the Siliguri Corridor — India’s core concern
North-central	495 sq km	Pasamlung and Jakarlung valleys	Pastoral and culturally significant to Bhutan; of limited strategic value to India
East	claimed from 2020	Sakteng sanctuary, Trashigang	A new claim, raised at a global environment forum, adjoining Arunachal Pradesh

• The Chinese offer has been consistent for decades: a package swap in which China recognises Bhutanese title to the larger north-central valleys in exchange for cession of the smaller but strategically decisive Doklam block.
• The arithmetic favours Bhutan and is adverse to India. Thimphu would gain area, settle its only territorial dispute and remove the threat of incursion; New Delhi would find Chinese positions overlooking the corridor.
• The eastern claim to Sakteng is read as leverage rather than a genuine historical assertion — a new claim created in order to be traded away.

The Negotiating Framework: 1988, 1998 and the Three-Step Roadmap

• Boundary talks began in 1984 and have run to 25 rounds, with a parallel series of expert group meetings.
• The 1988 agreement on guiding principles established that the boundary would be settled through peaceful negotiation on the basis of the existing situation.
• The 1998 agreement on peace and tranquillity committed both sides to refrain from unilateral action altering the status quo — the instrument invoked against Chinese road-building in 2017.
• The Three-Step Roadmap agreed in 2021 restructured the process into a sequence: agree the boundary on the table, conduct joint field visits, then formally delimit and demarcate.
  • The 15th Expert Group Meeting, held in Beijing at the turn of March and April 2026, established a Joint Technical Team on delimitation — the most concrete procedural step yet.
  • Bhutan and China have moved in parallel toward establishing diplomatic relations, which they have never had. These talks are advancing, not frozen.
• Bhutan’s motivation is sovereign, not pro-Chinese. A small state with an unsettled boundary against a great power is permanently exposed, and Doklam showed Thimphu that its unresolved border makes it a theatre for someone else’s rivalry.

Doklam 2017 and the Trijunction Dispute

• The standoff of June to August 2017 lasted 73 days, triggered by Chinese road construction on the Doklam plateau toward the Jampheri Ridge.
• India intervened on Bhutanese territory, which is legally unusual and rests on two grounds: the security relationship with Thimphu, and India’s own direct interest in the trijunction point.
• The legal core is where the trijunction lies.
  • China relies on the 1890 Anglo-Chinese Convention on Sikkim and Tibet, which defines the boundary as a watershed and which China reads as fixing the trijunction at Mount Gipmochi.
  • India’s position is that trijunction points cannot be fixed bilaterally and must be settled in consultation with the third state — an understanding reflected in a 2012 India–China agreement — and that the trijunction lies further north at Batang La.
  • Bhutan’s own position, stated publicly on 29 June 2017, was that the area is Bhutanese and that the construction violated the 1988 and 1998 agreements.
• The standoff ended in disengagement without resolution. Chinese activity in the wider Amo Chu valley has continued, including village and road construction in and around disputed tracts — a pattern documented by researchers including Robert Barnett.

Why a Bhutanese Boundary Is an Indian Security Question

• The Siliguri Corridor is the reason. It narrows to about 22 kilometres and is the only land connection between the Indian mainland and the eight north-eastern states, carrying rail, road, pipeline and power lines.
• The Chumbi Valley is a Chinese salient pointed at it, driving southward between Sikkim and Bhutan and shortened further by Chinese rail and road construction toward Yadong.
• Doklam is the shoulder of that salient. Chinese control of the plateau and of the Jampheri Ridge beyond would provide observation over the corridor and remove the depth on which Indian defensive planning depends.
• India’s structural advantage is that two of the triangle’s three arms rest on states aligned with it. A settlement transferring Doklam would convert a two-arm advantage into a one-arm advantage, which is the whole of India’s stake.
• The policy implication is where the scholarship divides.
  • S.D. Muni argues that normalisation is inevitable as Bhutanese democracy matures, that Indian attempts to prevent it will only confirm the image of an overbearing neighbour, and that India should secure its specific strategic concerns while accepting it.
  • Harsh V. Pant argues against any repetition of coercive pressure, as short-sighted, publicly damaging, and liable to turn Bhutan into another Nepal.
  • The convergent conclusion is that India cannot veto a Bhutanese settlement and should not try — its leverage lies in making the Doklam component costly to concede, and in deepening the economic relationship so that Bhutan’s stake in Indian goodwill exceeds its stake in a settled boundary.

The Bangladesh Boundary: The Model of What Settlement Looks Like

• This is the one South Asian land boundary dispute settled by agreement, and it is worth studying for what settlement actually required.
• The problem inherited in 1947 had three components.
  • Enclaves — sovereign fragments of one country entirely surrounded by the other, a legacy of pre-colonial arrangements between Cooch Behar and the Mughals. Some were counter-enclaves, and one was a counter-counter-enclave, the only one of its kind in the world.
  • Adverse possessions — tracts held and administered by one country but legally belonging to the other, covering some 7,000 acres.
  • Un-demarcated stretches — about 6.1 kilometres in three sectors where the line had never been fixed on the ground.
• Enclave residents were stateless in fact if not in law. They could not lawfully leave without transiting foreign territory, and so had no practical access to schools, hospitals, police or courts of either state.
  • Willem van Schendel’s work on the Bengal borderland and Brendan Whyte’s enclave scholarship documented the resulting deprivation in detail.

From 1974 to 2015

• The Indira–Mujib Agreement of 16 May 1974 set out the exchange of enclaves, the resolution of adverse possessions and the demarcation of the remaining stretches.
  • Bangladesh ratified it in 1974. India did not, because Indian ratification required something Bangladesh’s did not.
• The constitutional obstacle was decisive, and it is the article’s most transferable legal point.
  • In the Berubari Union reference, the Supreme Court held that cession of Indian territory to a foreign state cannot be effected by executive agreement and requires a constitutional amendment under Article 368, because it alters the territory of India defined in Article 1.
  • In Maganbhai Ishwarbhai Patel, decided after the Rann of Kutch award, the Court drew the crucial distinction: settling or ascertaining a disputed boundary is not cession and needs no amendment; transferring territory indisputably India’s does.
  • The enclaves were unambiguously Indian territory, so the 1974 agreement fell on the Berubari side of the line.
• The 100th Constitutional Amendment of 2015 supplied it, amending the First Schedule in respect of Assam, West Bengal, Meghalaya and Tripura to give effect to the 1974 agreement and its 2011 protocol.
  • It passed both Houses without a single dissenting vote — a level of consensus no other territorial question in India has commanded.
• The exchange took effect at midnight on 31 July 2015.
  • 111 Indian enclaves covering about 17,000 acres passed to Bangladesh; 51 Bangladeshi enclaves covering about 7,000 acres passed to India.
  • Residents were given a choice of nationality. Most in each set stayed where they were and took the surrounding state’s citizenship; fewer than a thousand people moved to India.
  • The operation was conducted without violence, without displacement on any scale and without litigation — which, for a transfer of sovereignty over some 50,000 people, is remarkable.

The 2014 Maritime Award: India Lost and Accepted

• The most useful single fact about Indian boundary policy is that India lost a boundary arbitration on the merits and complied with the result.
  • Bangladesh instituted proceedings under Annex VII of the Law of the Sea Convention in October 2009, before a tribunal at the Permanent Court of Arbitration presided over by Rüdiger Wolfrum, with P.S. Rao as India’s appointee.
  • The award of 7 July 2014 delimited the territorial sea, the exclusive economic zone and the continental shelf, including beyond 200 nautical miles.
  • Of roughly 25,600 square kilometres in dispute, the tribunal awarded Bangladesh about 19,500 — the substantial bulk of the claim.
• The reasoning was the standard method, and Bangladesh won on geography.
  • A provisional equidistance line was adjusted because the concavity of the Bangladeshi coast would otherwise have cut off its seaward projection between India and Myanmar.
  • The same reasoning had already succeeded against Myanmar before the International Tribunal for the Law of the Sea in 2012, and the two awards together settled the whole Bay of Bengal.
  • It produced a “grey area” in which Bangladesh holds continental shelf rights beneath waters where India holds exclusive economic zone rights — a solution requiring cooperation rather than a clean division.
  • It disposed of the New Moore / South Talpatti question — a live sovereignty dispute until rising sea level removed the island itself.
• India welcomed the award publicly and complied. That act is the strongest answer to the charge that India refuses third-party settlement of its boundaries — and the 2019 declaration to the World Court is the strongest evidence that it does not mean to repeat the experience.

India has settled every boundary it was willing to constitutionalise, and constitutionalised only the boundary where the domestic politics permitted it.

The Maritime Boundaries: Sri Lanka, Myanmar and the Wider Ocean

Kachchatheevu and the 1974–1976 Agreements

• Kachchatheevu is an uninhabited island of about 285 acres in the Palk Strait, of volcanic origin, carrying St. Anthony’s Church and an annual festival attended by fishermen of both countries.
• Two agreements fixed the boundary.
  • The maritime boundary agreement of 26 June 1974, concluded between Indira Gandhi and Sirimavo Bandaranaike, placed the island on the Sri Lankan side of the line in the Palk Strait.
  • The 1976 agreement extended the delimitation into the Gulf of Mannar and the Bay of Bengal and barred each country’s fishermen from the other’s exclusive economic zone.
• The controversy is whether the island was “ceded”, and the legal answer is precise.
  • If it was Indian territory, cession required a constitutional amendment under the Berubari rule; none was passed.
  • The Government of India’s consistent position has been that no Indian territory was ceded — that sovereignty over the island was disputed, that the 1974 agreement settled a disputed title, and that under Maganbhai the settlement of a disputed boundary requires no amendment.
  • The Tamil Nadu Assembly has passed resolutions demanding retrieval, petitions challenging the agreements have been pending in the Supreme Court since 2008, and a law officer of the Union told the Court that retrieval would require war.
• The living grievance is not sovereignty but fishing. The 1974 agreement preserved access for Indian fishermen to dry nets and visit the shrine without travel documents, but not to fish.
  • The Palk Bay fisheries dispute — driven by bottom trawling from the Indian side, enforced by arrests and boat seizures from the Sri Lankan side — has continued regardless of where the line runs.
• The case shows the central asymmetry in miniature. The boundary was settled cleanly and permanently in 1974; the resource conflict it was meant to resolve has never been settled at all.

Myanmar: The Under-Used Counter-Example

• India and Myanmar settled both boundaries, and nobody discusses it.
  • The land boundary was settled by the boundary agreement of 1967 and demarcated on the ground.
  • The maritime agreement of December 1986 covered the Andaman Sea, the Coco Channel and the Bay of Bengal — the first maritime delimitation concluded in the Bay, and a template for what followed.
  • The India–Myanmar–Thailand trijunction point was agreed separately in 1993.
• The instructive feature is the political context. The settlement was reached with a military government during a period of minimal warmth, which suggests that boundary settlement does not require good relations — only low salience and a tractable question.

The Pattern Across the Indian Ocean

• India has maritime agreements with every neighbour except Pakistan — Indonesia in 1974 and 1977, Sri Lanka in 1974 and 1976, the Maldives in 1976, Thailand in 1978 and Myanmar in 1986, along with trijunction agreements.
• The one exception proves the argument. The India–Pakistan maritime boundary is unsettled for a single reason — it cannot be drawn until Sir Creek is settled, because the creek’s terminus is the point from which the line must start.
• The remaining boundary, with Bangladesh, was settled by an adverse award India accepted. The complete Indian maritime picture is therefore: every boundary settled, one by litigation, one hostage to a marsh.

The Durand Line: A Boundary India Watches Rather Than Owns

• The Durand Line of 1893 divides Afghanistan from what is now Pakistan across some 2,600 kilometres, and Afghanistan has never accepted it.
  • It was fixed by Mortimer Durand, foreign secretary of British India, and Amir Abdur Rahman Khan, and runs through Pashtun tribal territory, dividing the same communities on either side.
  • Afghanistan’s position has been consistent across every regime — monarchy, republic, communist, mujahideen and both Taliban emirates — that the agreement was extracted under duress, limited in duration, or lapsed with British India.
  • Afghanistan was the only state to oppose Pakistan’s admission to the United Nations in 1947, and a loya jirga in 1949 declared the line void.
• Pakistan’s response has been to make the line a fact. It has fenced the great bulk of the alignment, which the Taliban government has protested and in places physically removed, producing armed clashes through 2025 and into 2026.
• The dispute is the origin of “strategic depth”. An unsettled western boundary and an irredentist Pashtunistan question gave Pakistan a permanent interest in a compliant Kabul, and therefore in the proxies it later turned eastward.
• India’s interest is derivative but real. This is the one boundary in the region whose non-settlement imposes costs on Pakistan rather than India, and the deterioration of Pakistan–Afghanistan relations since 2025 has driven India’s own re-engagement with Kabul.

Why India Settled Its Maritime Boundaries and Not Its Land Ones

This is the article’s central comparison, and the explanation is not that Indian diplomats tried harder at sea. It is that the two categories differ on almost every variable determining whether a dispute can be resolved, and the sea happened to be favourable on all of them at once.

Variable	Maritime boundaries	Land boundaries
Legal regime	Codified — UNCLOS and a settled three-stage method	Customary and fragmentary — competing principles, no code
Adjudicative forum	Available — ITLOS, PCA, ICJ, with compulsory procedures	None in practice; the 2019 declaration excludes boundaries
Population	None — nobody lives on a seabed	Settled populations with citizenship, land title and votes
Stakes	Divisible — area, fish, hydrocarbons, all quantifiable	Indivisible — sovereignty is not a quantity that can be split
Domestic salience	Very low — the 2014 award produced no protest	Very high — every metre is a national question
Constitutional entrenchment	None; delimitation is an executive act	Berubari: cession needs a constitutional amendment
Symbolism	Technical and invisible	Identity-defining — Tawang, Kashmir, Kalapani
Cost of concession	Reputationally negligible	Electorally lethal
Outcome	All settled except the one dependent on a land dispute	Two of six settled

• The decisive variable is population. A maritime boundary can be moved without moving anybody; a land boundary cannot. The moment human beings, land titles, revenue records and electoral rolls sit on the ground in question, the negotiation stops being about area and becomes about citizenship.
• The second is the forum. The Law of the Sea Convention created compulsory procedures with binding outcomes, so a party could be brought to a decision without consenting to that particular case. Nothing equivalent exists for land boundaries, and India’s 2019 reservations make certain nothing will.
• The third is salience. The 2014 award cost India some nineteen thousand square kilometres and generated no political controversy whatever, because almost no voter knew it had happened. No land concession of any size has ever been available on those terms.

The sea was settled because losing there was invisible; the land is unsettled because losing there is visible, and the visibility is the whole of the difficulty.

What Settlement Required Where It Happened

• Political capital spent at the very top. The Bangladesh settlement moved only when a Prime Minister chose to expend capital on it personally, against domestic opposition and with no electoral return.
• Cross-party consensus in constitutional form. The 100th Amendment passed unopposed, which means the opposition of the day chose not to make a territorial concession an issue — restraint without which the amendment was impossible.
• A state government’s veto had to be overcome. The Land Boundary Agreement was blocked for years by West Bengal’s objections and moved only when the state’s assent was obtained. Federalism is a real constraint — the boundary runs through states, not through Delhi.
• Acceptance of an adverse legal outcome. India complied with an award it had lost, and gained more from the compliance — the demonstration that it is a law-abiding maritime power, at a moment when it was arguing exactly that about the South China Sea — than it lost in seabed.
• A trade of asymmetric value. India gave up more area than it received in the enclave exchange and gained a settled, demarcated, manageable border with its most important regional partner. That is the shape every remaining settlement would have to take.

The Cost of Non-Settlement

• Militarisation on a scale no region sustains. Two land boundaries are held by the army rather than a border-guarding force, and on one the terrain kills more soldiers than the enemy does.
• The foreclosure of connectivity. A disputed boundary cannot host a road, a rail line, a pipeline or a transmission corridor — which is why South Asia has the lowest intra-regional trade share of any major region, and why the boundary question and the trade question are one question.
• Every local incident becomes a national one. Where the line is agreed, a herder crossing it is a police matter; where it is not, the same act is a territorial assertion requiring a government response. Non-settlement removes the option of treating a small event as small.
• Domestic politics in the smaller states is distorted. Anti-India mobilisation is cheapest where a boundary grievance exists to mobilise around, and Nepali politics has repeatedly demonstrated the point.

“Nepal’s internal politics is divisive; relations with India are a highly political issue and India is a convenient scapegoat for Nepal.” — V.P. Haran

Managing What Cannot Be Settled

• Border management is a response to non-settlement, not a substitute for it, and it was largely built after the Kargil conflict.
  • The Kargil Review Committee and the Group of Ministers process produced the one-border-one-force principle and a Department of Border Management in the Home Ministry.
  • Forces are allocated by border — the Border Security Force on Pakistan and Bangladesh, the Indo-Tibetan Border Police on China, the Sashastra Seema Bal on Nepal and Bhutan, and the Assam Rifles on Myanmar.
• Fencing has become the default instrument, and its limits are visible.
  • On the Bangladesh border the government’s own figures show roughly 3,232 km of 4,096.7 km fenced, about 864 km unfenced and about 175 km assessed as not feasible.
  • Fencing the Myanmar border was approved after the announcement ending the Free Movement Regime in February 2024, and is contested by Mizoram and Nagaland because it divides the same communities the Durand Line divides.
• The two open borders are the counter-argument to fencing everything. India’s boundaries with Nepal and Bhutan are crossed without passport or visa, and they are the two across which India faces no organised hostility. Settlement, not fencing, produced that.
• Development on the line has become policy. The Vibrant Villages Programme, approved in 2023 with an outlay of about ₹4,800 crore for the northern border and extended in 2025 to the other land borders, treats depopulation of frontier villages as a security problem.
  • The implicit recognition is that effectivités are built by residents, not by patrols.

Conclusion

The border disputes of South Asia are not, for the most part, disputes about which state deserves a piece of ground. They are the unfinished business of a boundary-making process interrupted at the demarcation stage in 1947 and never resumed, conducted over lines drawn by administrators solving a different problem, who never expected the results to be litigated.

• The disputes are technically soluble. Each has an identifiable legal question with an available answer — which stream is the Kali, whether Sir Creek is navigable, where the watershed runs north of NJ9842.
• They are politically frozen for the same reason in every case. The concession required is small and the domestic cost of making it is not, because a boundary concession is read as a concession about the nation itself.
• The Bangladesh settlement shows the price — a constitutional amendment, unanimous parliamentary support, a state government’s assent, a Prime Minister’s political capital and, at sea, the acceptance of a defeat.
• The maritime record rebuts fatalism. India has settled every maritime boundary it could, including one it lost, and the only unsettled stretch of water it has is unsettled because of a marsh.
• The direction of travel is management, not settlement — patrolling understandings with China, ceasefire arrangements with Pakistan, working groups with Nepal, fences with Bangladesh and Myanmar. Management is cheaper, and it works, and it never ends.

The strategic judgment is unwelcome and straightforward. Every year a boundary remains unsettled makes it harder to settle, because populations, infrastructure, constitutional texts and constituencies accumulate on either side of a line meant to be provisional. Nepal has written its claim into its Constitution; India has written its reservations into its declaration to the World Court. Both acts were rational, and both made settlement dearer than it was the year before.