In 30 Seconds
Thirty seven articles into this series, its readers have encountered the era’s numbers in fragments, tolls per massacre, per operation, per register, and the fragments deserve assembly, because the aggregate is both the era’s summary and its most contested artifact. This article gathers what the record supports about the cost of Punjab’s dark years, states the ranges honestly, and explains why the ranges exist, since the explanation, as this series has repeatedly found, is itself part of the history.
The Official Aggregate
The most commonly cited official figures, drawn from Punjab Police and union home ministry compilations for the years 1981 to 1995, place the conflict’s total dead above twenty one thousand: civilians in the vicinity of twelve thousand, militants around eight thousand, and police and security personnel around seventeen to eighteen hundred. The distribution across time is as telling as the total: the annual toll, in the hundreds through the mid 1980s, exceeded five thousand in 1991 at the peak this series has described, and collapsed to double and then single digits after 1993. Whatever the aggregate’s defects, the curve’s shape, the long climb after 1984 and the cliff after 1992, is beyond dispute and underlies every interpretive article this series has published.
Why the Aggregate Undercounts
The official total’s defects are systematic, and this series has met each in situ. The militant column includes the staged encounters, executions recorded as battles, which inflates militants and erases civilians simultaneously. The disappeared appear in no column at all: the 2,097 verified cremations of the sampled Amritsar grounds, the thousands of case files the successor investigations assembled across other districts, and Khalra’s order of magnitude estimate of twenty five thousand statewide occupy the space between the official aggregate and the fuller truth, and the refusal to audit that space, the subject of this batch’s final article, means every honest total must be given as a range: from the official twenty one thousand toward figures a third or more higher. The November 1984 dead of Delhi and beyond, three thousand plus by the counts this series has cited, belong to the era’s ledger though not to Punjab’s field statistics, and the era’s excess deaths, the suicides of the aftermath, the addiction harvest of the widow colonies, belong to no statistics at all.
The Distribution of the Dying
Within the ranges, the distributions hold and matter. The dead were overwhelmingly Punjabi, and by every analysis a substantial majority of the militancy’s own victims were Sikhs, the fact this series has cited as the demand’s central refutation; the Hindu dead of the selections and bombings constitute the largest deliberate communal category; the geographic concentration in the Majha districts, Amritsar and Tarn Taran above all, ran several fold the state average; and the demographic concentration was crueller still, rural men between fifteen and thirty five, the Green Revolution’s grandsons, dying at rates that emptied age cohorts from the border villages. Punjab lost, in the era’s fifteen years, roughly one resident in a thousand to its violence, and the terror districts multiples of that.
What the Numbers Are For
This series assembles the numbers for the reason Khalra read the registers: because counting is the first form of accountability, and contested counts are where impunity lives. Every party to the era has preferred its own arithmetic, the state its clean columns, the movement its martyrologies, and the space between them has held, for four decades, the uncounted. The ranges this article has stated are not agnosticism; they are the current boundary of the knowable, drawn where official India stopped the audit. The final articles of this series concern the stopping, and what moving the boundary would require. The dead, in whatever column, are owed at minimum the figure.
