Punjab · 1978–1980 · Chapter 7

The appeal that was never filed

Thirteen Sikhs died in Amritsar in April 1978. The accused were acquitted in January 1980. The state’s own prosecutor recommended an appeal, a later High Court judgment records that the Chief Minister approved the recommendation, and no appeal was ever presented. This is an account of a decision trail that has not been produced.

The shortest true sentence about the 1978 Amritsar case is that no appeal was ever filed against the acquittal. Everything contested sits in the space between that sentence and the question of why.

Two answers have circulated for four decades and this article rejects both. The first is that the Chief Minister refused to appeal. The second, offered as a correction to the first, is that he approved the appeal and then his government fell. The first is contradicted by the surviving judicial record. The second is a compression that converts an internal approval into a completed authorisation and lets the fall of a ministry stand in for an explanation. The difference is not a quibble. It is the difference between an accusation, a defence, and an unanswered question about what happened to a file.

What the record supports is narrower and, read carefully, harder for everyone involved. A 1981 Full Bench judgment, delivered after the judges summoned and examined the Punjab government record, recorded that the Chief Minister endorsed the Home Department’s 15 January 1980 recommendation that an appeal be filed — that he, in the judgment’s phrase, “set his seal of approval thereon.” The same judgment records that no direction was then given to the Public Prosecutor to present the appeal. No appeal was filed while that government remained in office. The Full Bench held the statutory process incomplete for precisely that reason: the appeal had never been presented.

An internal approval of a departmental recommendation is not the statutory direction. It is not the filing. It is not the prosecution of an appeal. A government that approves a course and does not take the operative step has not done the thing it approved, and the gap between the two is where this article lives.

What the evidence is, and what kind of evidence it is

The controlling document is a judgment from 1981 about events in 1980. That distinction carries most of the weight in what follows.

The court had the government file in front of it. Judges who summon a state record and describe what they find are producing evidence of a high order — better than recollection, better than contemporaneous reporting, better than almost anything else likely to survive. But they are producing a reconstruction and a legal characterisation of an administrative file, written a year later, in litigation whose question was whether an earlier decision to appeal had become final before an appeal was presented. The judgment is authoritative about what the judges found. It is not the file, and it does not reproduce the words the Chief Minister wrote.

That is why this article writes the High Court later recorded and never he ordered or he unequivocally authorised. “Set his seal of approval thereon” paraphrases an administrative effect. The underlying notation could have read Approved — file appeal. It could have read Approved — obtain senior counsel, or Approved as proposed, or Consult outside counsel before final action. Those are four different instructions with four different consequences for what should have happened next, and the distance between them is the distance between a government that tried and a government that stalled. Nobody writing about this today has read that notation.

The sequence

Chronology written from memory is how arguments in this territory collapse. Each entry carries the kind of evidence behind it.

13 Apr 1978The Amritsar clash. Thirteen Sikhs died. [RF] The figure carried in the historical record, to be confirmed against the FIR and the trial judgment.
Before trialThe case — the principal accused and approximately sixty other defendants — is transferred from Punjab to the Court of Session at Karnal by order of the Supreme Court of India. [OR] A judicial transfer, not a Prime Minister’s act.
4 Jan 1980The Court of Session at Karnal acquits the accused. [OR]
13 Jan 1980The Special Public Prosecutor recommends filing an appeal. [OR] Recorded in the 1981 judgment.
15 Jan 1980The Punjab Home Department concurs. The recommendation is placed before the Chief Minister. [OR]
15 Jan 1980The High Court later recorded that the Chief Minister “set his seal of approval thereon,” and characterised the order as approving the filing of an appeal. [OR] The words written on the file have not been recovered.
After 15 JanNo formal direction is given to the Public Prosecutor to present the appeal. A proposal arises instead to engage outside eminent counsel. [PF] The judgment states expressly that the direction was not given.
Jan–Feb 1980No appeal is filed while the ministry remains in office. [PF]
16 Feb 1980The elected Punjab ministry is dismissed; President’s Rule follows. [OR] A separate event. It is not an explanation for the four weeks that preceded it.
15 Mar 1980Under President’s Rule a new Advocate General recommends against an appeal. [OR]
1 Apr 1980The administration under President’s Rule decides not to appeal. No appeal was ever presented, and the Full Bench later held the statutory process incomplete for that reason. [PF]

Read whole, the sequence has an obvious shape and a missing middle. Everything to 15 January moves briskly in one direction: prosecutor, department, Chief Minister, within forty-eight hours of each other, all pointing at an appeal. Everything from 16 February moves briskly in the other: new administration, new law officer, new advice, decision closed within six weeks. Between them sit roughly four weeks in which the operative step was available, was not taken, and was replaced by a proposal to look for better counsel.

Those four weeks are the subject of this article — not because a conclusion can be drawn from them, but because they are the only part of the story a document would settle, and the argument has been conducted for forty years by people who did not have that document either.

Why the outside-counsel proposal is the hinge

A government that has just approved an appeal in a case with roughly sixty acquitted defendants and thirteen dead has two obvious next moves. It can direct the Public Prosecutor to present the appeal, preserving the position, and retain senior counsel afterwards to argue it. Or it can look for senior counsel first and direct the filing once counsel is engaged.

The first is what a government does when it wants the appeal on the record. The second is what a government does when it wants the appeal argued well — and it is also, indistinguishably from outside, what a government does when it does not want the appeal and prefers the clock to decide. That ambiguity is genuine and it is not resolvable by reasoning. It is resolvable by reading the file.

So the questions worth asking are procedural rather than moral. Was outside counsel actually approached, and who? Is there correspondence, or only a proposal? Was a memorandum of appeal drafted — a despatch register would show it. Did the Chief Minister’s office follow up after 15 January, and in what terms? Did anyone note that a limitation period was running? Each has a documentary answer that either exists or does not, and the difference between diligence and delay is written in them.

What this article refuses to say

“He refused to appeal.” Contradicted by the judgment, which records his approval of the departmental recommendation. A claim repeated for forty years is still a claim, and this one now has a document against it.

“He approved the appeal and then his government fell.” Too favourable and too compressed. It treats an internal approval as a completed authorisation and offers the dismissal as a cause, when the dismissal came four weeks after the approval and nothing examined here connects them. Correcting a myth with a defence is not correcting it.

“The Prime Minister moved the case out of Punjab.” The transfer to Karnal was a Supreme Court order. Attributing a judicial act to a politician is the kind of error that discredits everything published beside it.

Any characterisation of what the notation said. The judgment paraphrases the effect. Until the notation is recovered its wording is unknown, and an unknown is not a blank to be filled with whichever reading suits the argument.

There is a fifth refusal that matters more than the other four. An acquittal is not proof that an event did not happen. Thirteen people died; a court found the case against the accused not proved to the criminal standard. Those are different propositions, and the failure to appeal touches only the second. Nothing here should be read as an opinion about the guilt of anyone who stood trial at Karnal.

What would weaken this account

A thesis published without its strongest counterargument is an advertisement. Four readings would materially change what is written above, each stated in its most persuasive form.

Counsel for the criticised. Four weeks is not long to retain senior counsel in a case of this weight, and a government in its final weeks does not know it is in its final weeks. If correspondence shows counsel actually approached, the delay reads as diligence and this article’s hinge collapses.

The historian who disagrees. The appeal was abandoned on 1 April, under President’s Rule, on the advice of a law officer that administration appointed. Concentrating on the four weeks before the dismissal may be looking in the wrong place; the operative decision was taken by people the Chief Minister did not appoint.

The advocate who says this understates. A government that approves an appeal and then does not issue the direction produces the same outcome as a refusal, and the families of thirteen dead are entitled to notice that the distinction is procedural. Precision that reads as exculpation is its own kind of failure.

The lawyer reading the judgment. A Full Bench holding that the statutory process was incomplete is a finding about the legal status of the appeal, not about anyone’s conduct. Building a conduct narrative on a procedural holding is the first thing an opponent would attack.

The fourth is the strongest, and it is why this article ends in a document request rather than a verdict.

The record request

Sixteen documents would close most of what is open, listed in the order in which they would matter, with the office most likely to hold each.

DocumentLikely custodian
The original file carrying the 15 January 1980 notation, and any handwritten marginal note on itPunjab Home Department; Punjab Archives
The Special Public Prosecutor’s opinion of 13 January 1980Punjab Home Department
The Home Department concurrence of 15 January 1980Punjab Home Department
The cabinet or ministerial circulation sheetPunjab Home Department
The proposal to obtain outside eminent counsel, and the names proposedPunjab Law Department
Correspondence attempting to retain that counselPunjab Law Department
Any instruction to the Advocate General or Public Prosecutor before February 1980Office of the Advocate General, Punjab
Any communication from the Chief Minister’s office directing delay or further consultationPunjab Home Department
The diary or despatch register showing whether an appeal package was preparedPunjab Home Department
Any draft memorandum of appealPunjab Law Department
The Home Department appeal registerPunjab Home Department
Any explanation produced after the ministry fell as to why the filing had not occurredPunjab Home Department; the Governor’s secretariat
The Lal Singh judgment text, for verbatim quotation and paragraph citationPunjab and Haryana High Court registry
The Supreme Court transfer orderSupreme Court of India registry
The Karnal trial judgmentCourt of Session, Karnal
The FIR and the casualty record for 13 April 1978Punjab Police; district administration, Amritsar

Two things about that list are worth saying out loud. It does not assert that any of these documents survive — a records request establishes existence, and none has been made. And the first item is worth more than the other fifteen combined: a single line of handwriting would convert the central question of this article from an argument into a fact.

What remains true regardless

Every reading canvassed above — the accusatory, the exculpatory, and the procedural one this article prefers — leaves the same five facts standing.

Thirteen Sikhs died in Amritsar in April 1978.

The accused were acquitted at Karnal on 4 January 1980.

The state’s own prosecutor recommended an appeal.

A later High Court judgment says that recommendation received the Chief Minister’s approval.

No appeal ever reached the High Court.

That is an institutional failure whoever caused it, and locating the individual responsible is a smaller question than it appears. A state prosecutes on behalf of the public. When its own prosecutor recommends an appeal in a case of thirteen deaths, the machinery that turns that recommendation into a filing is not discretionary in any ordinary sense — it is the ordinary functioning of the office. The machinery did not turn. What this article establishes is that it did not turn, and roughly when it stopped.

What it cannot establish, and will not pretend to, is who stopped it. Community memory has an answer, and it is not corrected by substituting an equally simplistic retrospective defence. It is corrected, if at all, by a file.

Read next: the Punjab desk and how to read it · editorial standards · methodology