The Sickle Changes Hands. The Grass Keeps Growing.
The Head Was Priced. The Body Was Denied. From Shahid Ganj to Bhai Jaswant Singh Ji Khalra
Plain-English renderings of Gurbani are mine. Gurmukhi remains primary.
This essay compares two very different periods. It does not claim that one institution or policy passed directly from the eighteenth century into the twentieth. The states, laws and formal target categories were different. Its question is narrower: what happens when the capture or killing of a targeted person becomes administratively valuable?
The head on the palm
Guru Nanak Sahib Ji says:
ਜਉ ਤਉ ਪ੍ਰੇਮ ਖੇਲਣ ਕਾ ਚਾਉ ॥
ਸਿਰੁ ਧਰਿ ਤਲੀ ਗਲੀ ਮੇਰੀ ਆਉ ॥
ਇਤੁ ਮਾਰਗਿ ਪੈਰੁ ਧਰੀਜੈ ॥
ਸਿਰੁ ਦੀਜੈ ਕਾਣਿ ਨ ਕੀਜੈ ॥੨੦॥jau tau prem khelan ka chau
sir dhar tali gali meri aau
it marag pair dharijai
sir dijai kan na kijaiAng 1412
Plain-English sense: If you long to play the game of love, come to my path with your head on your palm. Once you place your foot on this path, give your head and do not hold back. (SearchGurbani.com)
The head in these lines is not merely a future severed object. It carries the person’s thinking, judging, defending and claim to direct the path on their own terms. To place it on the palm is to surrender self-rule before entering the Guru’s way.
The line also gives us a guardrail for speaking about Shaheedi, usually translated as martyrdom. Shaheedi is not created simply because someone later dies violently. The path has already asked for the head. Bodily Shaheedi, where it is rightly named, is the outward completion of an inward giving that cannot be manufactured by the killer.
The history that follows concerns the opposite act.
The Guru asks for the head to be given in love.
The state makes the head payable in violence.
These are not the same act.
The first gives up the self’s claim to rule. The second claims power over someone else’s life: the power to locate, report, capture, kill, display, deny or enter that person into an account.
What follows is the story of two such accounts. The first was openly declared. The second had to be reconstructed from the paperwork left around bodies described as unidentified or unclaimed.
There is no demonstrated administrative line connecting them. The eighteenth-century system explicitly targeted the Sikh as Sikh. The modern system formally targeted named and alleged militants.
But in both periods, a targeted person’s death could become something claimable, rewardable and recordable.
The first ledger was open
The eighteenth-century reward schedule is known through historical writing and the earlier chronicles on which that writing relies. We do not have an original Lahore government proclamation before us. That source distinction matters.
In his account of events around 1740, Hari Ram Gupta reports that the government of Lahore under Zakariya Khan circulated a graded schedule of rewards against the Sikhs. Blankets and bedding were offered to anyone who cut a Sikh’s hair. Ten rupees were offered for information about Sikh whereabouts. Fifty rupees were offered to anyone who “caught or killed a Sikh.” The plunder of Sikh homes was permitted. Sheltering a Sikh or failing to report one could be punished by death; supplying food could lead to forced conversion. (Internet Archive)
The exact wording should be preserved. Gupta does not write “fifty rupees for a Sikh scalp” in this passage. He writes fifty rupees for catching or killing a Sikh. The shorter scalp formulation appears in later retellings, but it should not be placed inside Gupta’s mouth.
The structure is unmistakable.
Information had a price.
Capture had a price.
Killing had a price.
Shelter became an offence.
The state did not merely permit persecution. It tried to draw village officials, informers, neighbours and potential helpers into the same system of punishment and reward.
There was no single timeless price for the Sikh head. The rate varied across administrations and campaigns. During Lakhpat Rai’s campaign of 1746, for example, Gupta reports a payment of five rupees for every Sikh head. The changing amount is part of the history. What remained recognisable was the conversion of the Sikh person into a payable claim. (Internet Archive)
Shahid Ganj: the body as public instruction
Gupta records that captured Sikhs were brought in chains to Lahore and executed near the Nakhas outside the Delhi Gate, at the place later called Shahid Ganj. (Internet Archive)
The punishment did more than remove a person.
It instructed everyone who watched.
A public execution tells the population what the ruler believes it may do. The prisoner, the execution ground and the visible body become part of one message: this is what happens to the person placed outside protection.
The body becomes public instruction.
An earlier and separate episode shows this use of spectacle with particular force. After the capture of Banda Singh Bahadur and his companions in December 1715, the prisoners were taken first to Lahore and then to Delhi. At the Delhi procession in February 1716, Gupta reports that 2,000 Sikh heads were displayed on spears. (Internet Archive)
That episode belongs to the history of public display. It does not belong to Zakariya Khan’s later reward schedule as though the two were one continuous operation.
The pattern may be compared.
The events must remain distinct.
A true pattern is weakened, not strengthened, when separate events are forced into one scene.
The grass answered the sickle
The familiar sickle-and-grass verse belongs to the later persecution under Mu’in-ul-Mulk, remembered as Mir Mannu.
In a concise English rendering:
Mir Mannu is the sickle; we are the grass.
As he cuts us, we grow again.
This was not Mir Mannu’s boast.
Gupta says the Sikhs consoled themselves by singing a couplet “of their own composition.” The words belonged to the people being cut, not to the ruler doing the cutting. (Internet Archive)
That correction changes the meaning.
Had the verse been the ruler’s boast, the sickle would have owned the metaphor. Instead, the Sikhs placed the ruler’s violence inside another account.
The verse does not deny the cutting. It does not romanticise the dead or make bereavement harmless. It refuses the cutter’s claim to decide what the cutting will finally mean.
The sickle can cut what stands above the ground.
It cannot command the life from which the grass returns.
The ruler counted the severed.
The Panth named what the count could not finish.
Here, Panth means the Sikh collective living under the Guru’s path.
Two and a half centuries later, the hand had changed, the state had changed and the legal category had changed. Yet death could again become a rewarded claim, and the body could again enter an administrative account.
This time the account was not proclaimed at a market.
It had to be recovered from cremation registers and receipts for firewood.
The register that should not have survived
Bhai Jaswant Singh Ji Khalra did not begin with a mass grave.
He began with records.
Working with Jaspal Singh Dhillon Ji, he investigated bodies regularly brought by police personnel to municipal cremation grounds and declared unclaimed. Their press note of 16 January 1995 described the receipt books through which firewood was issued for the disposal of those bodies. The entries preserved dates and numbers even where the arrest, identity or death of the person had been denied or concealed. (Legal Tools)
That administrative detail was the centre of the method.
An arrest could be denied. A station diary could remain blank. A family could be told that its son had absconded or gone abroad.
But a body delivered to a cremation ground still had to be received. Wood still had to be issued. Someone might record a date, a number or a payment.
One state record could deny the person.
Another could preserve the body’s passage.
The body leaves a trace.
The figures that emerged count different things and must not be merged.
On 22 July 1996, a CBI interim report disclosed 984 illegal cremations at a crematorium in Tarn Taran, then within Amritsar district, between 1984 and 1994. The Supreme Court directed the CBI to continue its investigation and issue a public notice seeking assistance. (Reduced to Ashes)
The CBI’s later report listed 2,097 illegal cremations at three cremation grounds in Amritsar district, then one of Panjab’s thirteen districts. Human Rights Watch records that Bhai Khalra Ji had discussed more than 6,000 cremations in Amritsar district, a wider investigative figure than the official three-crematoria case set. (Human Rights Watch)
The National Human Rights Commission later acknowledged 38 duplicate entries, leaving 2,059 unique cremation records. (HRDAG and Ensaaf)
These are not four competing answers to one question.
The 984 figure was an interim Tarn Taran finding. The 2,097 figure was the original official three-crematoria case set. The 2,059 figure was that set after duplicate entries were removed. The figure above 6,000 belonged to the wider Amritsar-district investigation associated with Bhai Khalra Ji and Jaspal Singh Dhillon Ji.
None was a Panjab-wide total.
Keeping those distinctions visible does not diminish the finding. It prevents the dead from being turned into an impressive but indefensible statistic.
The second ledger used the language of law
The cremation records show where bodies passed. To understand one of the incentive systems operating around claimed encounters, we have to step back six years.
On 30 August 1989, the Director-General of Police, Punjab, issued an order to district superintendents offering financial rewards for the “apprehension/liquidation” of 53 named men described as “terrorists/extremists.” Amnesty International reproduced the order in its 1991 report. (Amnesty International, 1991)
The wording matters.
Apprehension belongs to arrest.
Liquidation placed killing inside the same reward instrument.
The later history must also be stated correctly. In April 1990, the Attorney General confirmed to the Supreme Court that the earlier order had existed but said it had lapsed. A replacement dated 1 April 1990 offered rewards for the “arrest/apprehension” of 41 men and removed the word “liquidation.” (Amnesty International, 1991)
The 1989 order should therefore not be described as though it remained formally operative without interruption.
But the wider incentive structure did not end with that one document. Amnesty International later reported that alleged militants were placed into categories, with rewards in the early 1990s ranging from approximately ₹50,000 to ₹500,000. It also reported that the central government had sanctioned a special fund, reportedly operated by senior police officers at state level, alongside alleged misappropriation of reward money and out-of-turn promotions for officers credited with killing large numbers of militants. These are attributed human-rights findings, not the results of a complete public audit. (Amnesty International)
One further figure is repeatedly quoted:
41,000.
It is often transformed into 41,000 extrajudicial executions.
That is not what the source establishes.
In its report for 1993, the United States Department of State recorded that the chief minister of Punjab told the state assembly that more than 41,000 bounties had been paid between 1991 and 1993. The same passage added that, in some cases, more than one person claimed credit for the same killing. (Refworld)
The attribution and the caveat belong together.
This was the chief minister’s statement to the assembly, as recorded by the Department of State. It was not an independently audited Department of State count.
Nor does it establish:
41,000 separate victims;
41,000 separate killings;
41,000 extrajudicial executions;
or 41,000 distinct encounters.
Because more than one person sometimes claimed credit for the same killing, the figure cannot be read as a count of distinct killings or victims.
The 41,000 figure is an attributed count of bounty payments.
It is not a count of victims.
To inflate the number would not honour the dead. It would place a weak claim where a strong and defensible one already exists.
The difference between the two historical periods must also remain visible. The eighteenth-century schedule, as reported by Gupta, explicitly targeted the Sikh as Sikh. The 1989 order named particular people and classified them as terrorists or extremists.
But the modern label was not self-proving. Human-rights investigators documented arbitrary detention, enforced disappearance and extrajudicial execution affecting alleged militants, suspected supporters, civilians and people already in police custody. Human Rights Watch concluded that the reward and incentive system contributed to increased disappearances and extrajudicial executions of civilians and militants alike. (Human Rights Watch)
The careful comparison is therefore this:
The eighteenth-century state formally targeted the Sikh as Sikh.
The modern state formally targeted named and alleged militants within a counter-insurgency whose categories did not remain confined to proven combatants.
Sikh militant groups also committed grave abuses. Human Rights Watch documented massacres of civilians, attacks on Hindu minorities, indiscriminate bomb attacks in crowded places, assassinations of political leaders, widespread extortion and land grabs. That record belongs in any truthful account of the decade. (Human Rights Watch)
But fairness does not require false equivalence.
The state controls lawful arrest, custody, station diaries, inquests, post-mortems, official force and the disposal of bodies. Documenting militant crimes does not answer what the state did to a person after arrest or while under its control.
Nor may we write a perfect mechanical chain from every bounty to every encounter and from every encounter to every cremation.
A cremation entry does not by itself prove that a killing was committed for money.
A bounty payment does not identify a particular body.
A police description of an encounter does not, by itself, prove either that the encounter was genuine or that it was staged.
Each individual chain must be established from its own evidence.
The stronger conclusion is structural: the reward system made the claimed capture or killing of an alleged militant administratively valuable, while unrecorded custody and secret disposal could remove the person and the evidence from independent scrutiny.
The two ledgers could reinforce one another.
They must not be collapsed into one.
The result was shown; the person was hidden
The difference between the two periods is sometimes expressed simply:
the eighteenth-century state displayed the head;
the twentieth-century state hid the body.
That is close, but incomplete.
The modern apparatus also displayed something. Encounter totals could be announced as evidence that the counter-insurgency was succeeding. A claimed killing could be treated as an operational achievement and could bring a bounty or professional advancement. (Refworld; Amnesty International; Human Rights Watch)
What could remain hidden was the path leading to that result:
the earlier arrest;
the unrecorded detention;
the fact of custody;
the person’s real identity;
the evidence surrounding the shooting;
and the location or disposal of the remains.
The older regime displayed the punished Sikh.
The modern apparatus could display the “encounter victory” while making the person behind it deniable.
The movement was therefore not simply from display to disappearance. It was from public punishment meant to be seen to official procedure capable of concealing responsibility.
In compressed language, it was a movement from sovereign spectacle to bureaucratic deniability.
The first message was:
See what has been done to the Sikh.
The second could become:
See the militant we have eliminated. Do not ask how this person entered our hands, what happened before the shooting, or where the body went afterwards.
The state still displayed a result.
It changed what the public was permitted to see.
The man who opened the register was made to disappear
Punjab Police abducted Bhai Jaswant Singh Ji Khalra from outside his home on 6 September 1995.
The evidence accepted by the courts established that he was taken to Police Station Jhabal, secretly detained and beaten. A witness testified that Bhai Khalra Ji was shot, that his body was transported to the Harike area, and that it was thrown into a canal. His body was never recovered. (Indian Kanoon)
The person who had made unclaimed bodies visible was himself denied an ordinary public record.
His arrest was denied.
His detention was concealed.
His body was not returned.
But the attempt to remove him from law did not succeed completely. Bibi Paramjit Kaur Khalra Ji petitioned the Supreme Court, which transferred the investigation to the CBI. The proceedings were long, but convictions were eventually obtained. (Indian Kanoon)
The legal sequence should be stated precisely.
On 18 November 2005, six Punjab Police officers were convicted. Jaspal Singh and Amarjit Singh were sentenced to life imprisonment for murder under sections 302/34 of the Indian Penal Code, with further sentences for criminal conspiracy, abduction with intent to murder and destruction of evidence. Prithipal Singh, Satnam Singh, Surinderpal Singh and Jasbir Singh received five years for criminal conspiracy and seven years for abduction with intent to murder. (Indian Kanoon)
On 8 October 2007, the Punjab and Haryana High Court acquitted Amarjit Singh while maintaining the convictions of the other five officers.
On 16 October 2007, following Bibi Paramjit Kaur Khalra Ji’s revision petition, the High Court enhanced the sentences of Satnam Singh, Surinderpal Singh, Jasbir Singh and Prithipal Singh for abduction with intent to murder from seven years to life imprisonment. Jaspal Singh’s existing life sentence remained in place. (Indian Kanoon)
On 4 November 2011, the Supreme Court dismissed the appeals of the five remaining convicted officers. Their convictions and life sentences stood. (Indian Kanoon)
That accountability was real.
It was also incomplete.
The convictions established individual criminal responsibility for the abduction and murder of Bhai Khalra Ji. They did not identify every person in the cremation records, reconstruct every bounty payment, or open the full police, medical, municipal and governmental account across Panjab.
The register had been opened.
The full account remained unfinished.
The archive and the Ardaas
The archive can establish that an order existed.
It can establish that a reward was offered, that a body passed through a cremation ground, that wood was purchased, that custody was denied, or that a crime was committed. It can establish individual responsibility where the evidence permits.
That work cannot be replaced by general remembrance.
A family may know that someone was taken. A community may know that an official explanation is false. But public truth also requires the date, the entry, the signature, the medical record, the court finding and the responsible office.
The archive answers denial.
It does not answer every question.
A cremation register can establish that a body was burned. It cannot, by itself, certify Shaheedi.
A judgment can establish murder. It cannot see the whole inward condition of the person murdered.
The police word terrorist cannot be accepted merely because an officer wrote it. But the word Shaheed should not become an automatic counter-label applied to every person killed by the state.
That distinction protects the dead in both directions.
It refuses the state’s unproven category.
It also refuses to make Sikh remembrance another careless system of classification.
The records establish what happened.
Shabad Guru governs what Shaheedi means.
Ardaas is the Sikh prayer of remembrance, supplication and standing before the Guru.
It is not an archive. It does not replace an arrest register, an inquest, a post-mortem, a cremation record, a court judgment or a public inquiry. To invoke Ardaas as a reason not to retrieve the file would be a failure of seva.
The archive and the Ardaas answer different dangers.
The archive answers the danger of denial. It says: this person existed; this officer signed; this body passed here; this order was issued; this crime occurred.
Ardaas answers the attempt by power to control the final meaning of what it did.
The eighteenth-century regime displayed the Sikh body so that the population would remember the person as a defeated object of the ruler.
The modern apparatus could hide the person behind words such as unidentified, unclaimed, absconding or gone abroad.
Display and disappearance appear to be opposites.
Both seek authority over memory.
The first commands:
Remember this person only as what the ruler destroyed.
The second commands:
Do not remember this person at all.
Ardaas refuses both instructions.
But Ardaas does not float free as folklore. The Panth does not stand beneath the sickle-and-grass verse, the nation, the state, the archive or the author.
The Panth stands beneath Shabad Guru Granth Sahib Ji.
Guru Nanak Sahib Ji gives the order:
ਸਬਦੁ ਗੁਰੂ ਸੁਰਤਿ ਧੁਨਿ ਚੇਲਾ ॥
sabad guroo surat dhun chelaa
Ang 943
Plain-English sense: Shabad is Guru; the attuned consciousness is the disciple. (SearchGurbani.com)
The sickle-and-grass couplet carries a people’s answer to persecution.
The cremation register carries the trace of a hidden body.
The judgment carries the finding of a crime.
None stands above the Guru.
Ardaas functions as a daily Panthic register of remembrance spoken beneath that authority. It does not preserve every individual name or file number. That is why the documentary work remains necessary. But it refuses to let the state’s category become the final account of the person.
The archive counts so the state cannot deny.
Ardaas remembers under the Guru so the state cannot define.
The grass does not erase the cut.
It denies the sickle the last word.
Bhai Jaswant Singh Ji Khalra took a cremation entry and turned it back into a question. He turned that question towards the state. He turned the unclaimed body back towards the family. He turned the count back towards the person.
The sickle did not receive the final word.
The fire did not receive the final word.
The state entered a price.
It did not close the account.
Verify
Opening Gurbani: ਜਉ ਤਉ ਪ੍ਰੇਮ ਖੇਲਣ ਕਾ ਚਾਉ, Salok Vaaran Te Vadheek, Slok 20, Mahala 1, Ang 1412. Gurmukhi and Ang were cross-checked against SearchGurbani and SriGranth. (SearchGurbani.com; SriGranth.org)
Eighteenth-century schedule: Hari Ram Gupta, History of the Sikhs, vol. II, pp. 57–58, reports blankets and bedding for cutting a Sikh’s hair, ten rupees for information, and fifty rupees for catching or killing a Sikh. His account is a historical synthesis, not a surviving original proclamation. (Internet Archive)
Shahid Ganj, the head-price and the folk verse: Gupta records the executions near the Nakhas at p. 56, the five-rupee head-price during Lakhpat Rai’s 1746 campaign at p. 77, and the sickle-and-grass couplet as the Sikhs’ own composition at p. 115. (Internet Archive)
The Banda Singh Bahadur procession: Gupta’s account of the displayed-head procession towards Delhi appears at pp. 30–31 and belongs to the separate 1715–16 episode, not to Zakariya Khan’s later reward schedule. (Internet Archive)
Cremation records: The 16 January 1995 press note records the use of municipal firewood receipt books. (Legal Tools) Reduced to Ashes (Ram Narayan Kumar, Amrik Singh, Ashok Agrwaal and Jaskaran Kaur, South Asia Forum for Human Rights, Kathmandu, 2003) dates the 984-cremation interim report to 22 July 1996 and records the Supreme Court’s direction to continue the investigation and issue a public notice. (Reduced to Ashes)
Cremation totals: Human Rights Watch records the 2,097-entry three-crematoria case set and the wider figure above 6,000. Ensaaf and HRDAG record the NHRC’s acknowledgement of 38 duplicate entries, leaving 2,059 unique cremation records. (Human Rights Watch; HRDAG and Ensaaf)
Police reward orders: Amnesty International’s 1991 report reproduces the 30 August 1989 “apprehension/liquidation” order for 53 men and records the 1 April 1990 “arrest/apprehension” replacement for 41 men. (Amnesty International, 1991)
The 41,000 figure: The source is the chief minister of Punjab’s statement to the state assembly, recorded by the United States Department of State, with the express caveat that more than one person sometimes claimed credit for the same killing. (Refworld)
Bhai Khalra Ji’s case: The abduction, illegal detention, murder and disposal of the body, together with the trial convictions of 18 November 2005, the High Court decisions of 8 and 16 October 2007, and the Supreme Court decision of 4 November 2011, were checked against the final judgment. (Indian Kanoon)
Closing Gurbani: ਸਬਦੁ ਗੁਰੂ ਸੁਰਤਿ ਧੁਨਿ ਚੇਲਾ, Raamkali, Sidh Gosht, Mahala 1, Ang 943. (SearchGurbani.com; SriGranth.org)
Cross-check
Every Gurbani line should be checked directly against Shabad Guru Granth Sahib Ji. Gurmukhi remains primary; Romanisation and English are learning aids.
The eighteenth-century claims should be checked against the relevant edition of Hari Ram Gupta and, for formal historical scholarship, against the earlier chronicles he cites. The modern reward orders should be checked against Amnesty International’s 1991 reproduction. The 41,000 statement should always retain both its attribution and its multiple-claim caveat. The cremation figures must remain attached to their particular geography, record set and stage of inquiry. The conviction chronology should be checked against the Supreme Court judgment.
Where a source reports a statement, this piece reports it as a statement. It does not silently convert an attributed claim into an independently audited finding.
Correction note
This version makes three corrections visible.
Catch or kill, not Gupta’s “scalp”: Gupta’s cited passage says fifty rupees for anyone who “caught or killed a Sikh.” The shorter scalp formulation appears in later retellings but is not his wording there.
The couplet belonged to the Sikhs: The sickle-and-grass verse was not Mir Mannu’s boast. Gupta describes it as the Sikhs’ own composition.
The Banda Singh Bahadur procession was separate: The procession of prisoners and displayed heads towards Delhi belongs to the 1715–16 Banda Singh Bahadur episode. It is relevant to the history of public spectacle but is not evidence of Zakariya Khan’s later reward schedule.
A piece that asks governments to preserve and correct their records must correct its own publicly, calmly and with a dated note.
Source note
The eighteenth-century schedule is historiographically attested through Hari Ram Gupta’s synthesis. It is not presented as a surviving original proclamation.
The 30 August 1989 reward order is directly documented through Amnesty International’s reproduction.
The ₹50,000 to ₹500,000 range, reported special fund, alleged misappropriation and promotion system are attributed human-rights findings, not a complete public audit.
The 41,000 figure is an attributed official statement about bounty payments, not a verified victim count.
The 984, 2,097, 2,059 and above-6,000 cremation figures belong to different record sets and stages of inquiry. They are not combined.
No one-to-one causal link is assumed between every bounty payment, every claimed encounter and every cremation entry. Such links must be established individually.
The main title, “The Sickle Changes Hands. The Grass Keeps Growing,” is the author’s echo of the historical couplet. It is not presented as the wording of the original verse, which names Mir Mannu.
“Panjab” is used in the essay’s prose. Official names retain their legal spelling, including Punjab Police, Director-General of Police, Punjab, chief minister of Punjab, State of Punjab, and Punjab and Haryana High Court.
The distinction between public punishment and official deniability, and between what the archive establishes and what Ardaas remembers under the Guru, is the author’s analysis.
Strong disagreement is welcome.
Contempt is not.
Bhul chuk maaf karni Ji.
Gurjit Singh Sandhu
PanthSeva
London