Genocide in Gujarat: Patterns of violence

Concerned Citizens Tribunal ­ Gujarat 2002

The Concerned Citizens Tribunal ­ Gujarat 2002, was conceived as a response to the carnage that rocked the state of Gujarat following the Godhra tragedy [when the partial incineration of a train was blamed on Islamic militants as a pretext for the massacre that followed] on February 27, 2002. The eight-member Tribunal was constituted in consultation with a large number of groups from within Gujarat and the rest of the country.

The Tribunal collected 2,094 oral and written testimonies, both individual and collective, from victim-survivors and also independent human rights groups, women’s groups, NGOs and academics. The documentation work done by relief camp managers and community leaders, from lists of persons killed or ‘missing’, to the meticulous tabulation of economic loss and religious desecration, is unprecedented and immense.

The Tribunal pays tribute to the victim-survivors, individually and collectively, who deposed before us at great risk to their person in the simple hope that one day justice will be done and the guilty be punished. Even as the Tribunal sat in Ahmedabad, there were threats and premises like the circuit house at Shahibaug were denied us due to the omnipresence of prowling mobs. We acknowledge our great debt to the activists on the ground who worked day and night to bring the victims and reliable eyewitnesses to us.

After recording evidence, visiting sites, placing on record statements and collecting other relevant material, the Tribunal arrived at some prima facie conclusions. These were forwarded along with our recommendations to both the central and state governments and their views were awaited. However, the Tribunal regrets that neither the state government nor the central government, or individual ministers to whom request letters were sent, responded. Though we are entitled to draw adverse conclusions from this lack of response, because that they did not respond to the interim findings, we do not propose to do so.

In a democracy, the people’s right to information should be paramount. Any government wedded to this basic right should have eagerly cooperated in the effort of a citizen’s tribunal to inquire and let the people know what happened in Gujarat, who engineered the carnage, and who the guilty are.

The Tribunal undertook this huge task as part of the exercise of this fundamental human freedom. The health of any society lies not in denials and half-truths when grave injustices have occurred, but in courageously admitting to them, righting those wrongs with justice and then reconciliation. That both the government of Gujarat and the government of India did not participate in the inquiry reveals their utter disregard for the people’s basic democratic right to know.

Having completed its task, it is with humility that the Tribunal presents this report to the country and the world. Even as we complete our task, we know and recognise that our country’s record in the matter of punishment of the guilty in cases of mass crimes, against the minorities, against Dalits, and against the poor has been pathetic. Yet, with hope that is eternal to the human condition, we do present this report in the belief that, this time, knowing the truth will help us chalk a future that is radically different.

The sorry state of the rule of law in the country is closely connected to the functioning and accountability of our courts, and the criminal justice system is crying out for radical reform. We hope that with justice to the victim-survivors, these reforms will become a matter of urgent political debate.

The panel that constituted the Tribunal pays a humble tribute to all the hapless and innocent victims of the ghastly Gujarat carnage. We dedicate this report to them and to their surviving relatives. And also to each one of those women and men who, at great risk to their person, provided succour and helped expose the truth.

Patterns of violence 

A noticeable feature of the Gujarat carnage is the distinct and similar patterns that have emerged from different parts of the state. While some local conditions and socio-economic factors do differentiate the attacks from one place to another, detailed and extensive evidence before the Tribunal points to the overwhelming and sinister similarity of the attacks that were engineered and launched. This is evident in the manner in which innocent people were quarterised, sometimes sexually violated and killed; in the ammunition used for the gory killings and the arson; in the immediate and long-term preparations for the violence. All these are detailed below.

Selective targeting of Muslims 

From the extensive evidence recorded by the Tribunal, it is clear that Muslims from all social strata, rich and poor, were the prime targets for the state-sponsored pogrom unleashed all over the state of Gujarat. From cities and towns to villages, be it the question of life, dignity or property, barring few exceptions, Muslims were the sole target. While the targeting of economically better off Muslims was limited to their property, and this damage was vast and extensive, the lower middle class and the working class sector, be it in urban centres or villages, faced attacks on their life, property and dignity. Except in the few cases where some Hindu establishments were targeted (in the immediate vicinity of areas that have been converted into Muslim ghettos), in cities like Ahmedabad and Vadodara, the recent carnage was marked (unlike earlier rounds of violence where sections of both communities were affected) by the selective targeting of Muslim lives, Muslim homes, Muslim business establishments and Muslim properties. Whether it was on the posh CG Road of Ahmedabad, the main streets of Bharuch, Ankleshwar and Vadodara, or the villages of Kheda district or the Panchmahal, small and large farms and properties, homes and shops, only of Muslims were the target of marauding mobs. A potentially gruesome tragedy, where the rampaging mobs nearly set upon and burnt alive 70 children in a Muslim-run orphanage in the city of Bhavnagar, was averted by a conscientious police official loyal to his uniform, is worthy of mention here. For having shown exemplary courage and saving innocent lives, the SP Bhavnagar, Rahul Sharma was ‘rewarded’ with a summary transfer.

In most places, Hindu houses amongst Muslim bastis had been marked out before the attacks using saffron flags, or pictures of Ram and Hanuman, or with crosses. Evidence before the Tribunal shows that in some places this marking was done a few days before February 27 and which was the ostensible justification for the ‘retaliation’. These markings were to avoid inadvertent attacks on Hindu homes and businesses in areas that were targeted later.

There was no damage whatsoever to the Hindu houses so marked. Months later, saffron flags were still aflutter in many villages of Gujarat and it is evident how the attacks and destruction were carried out so that the Hindu houses were not damaged. In some villages, the adjoining Hindu houses were first sawed away from the Muslim houses before the latter were set on fire. Each attack, therefore, took not just extensive planning but also several hours to execute, which further indicates an abdication of responsibility by the police in its failure to come to the rescue of the targeted community.

From the state wide evidence earlier recorded and placed before us, it is also clear that apart from the lives of Muslims, several symbols of India’s composite culture were deliberate targets during the carnage in Gujarat. The durgahs (shrines) of Sufi saints that are revered by persons from all communities, especially the oppressed castes, deserve special mention here.

The other targets of violence were couples who had entered into inter-community marriages. Evidence was specifically placed before us about the shameful stripping, gross sexual abuse and subsequent quartering and killing of Geeta (Mumtazbano), a Hindu woman from Ahmedabad who had married a Muslim man, Salim. The couple was tricked into visiting her family on April 5. They were set upon while travelling on a scooter. Geeta died while Salim survived.

Violence against mixed couples has become common all over Gujarat and the issue of inter-religious marriage has become part of the hate propaganda against Muslims and those Hindus who enter into or accept such marriages.

Brutality and bestiality of attacks 

The widespread violence that targeted Muslims in urban and rural Gujarat was marked by utter bestiality and brutality. We have recorded evidence from Naroda Patiya in Ahmedabad, as also from witnesses from Kheda, Bharuch, Ankleshwar, Panchmahal, Mehsana, Sabarkantha, Banaskantha and Vadodara, that training camps were conducted by the Bajrang Dal and the VHP [Vishna Hindu Parishad], backed by the RSS [Rashtriya Swayamsevak Sangh] and supported by democratically elected representatives from the ruling BJP [Bharatiya Janata Party]. The camps were often conducted in temples. The aim was to generate intense hatred against Muslims painted as ‘the enemy’, because of which violence was both glorified through the distribution of trishuls and swords, and justified as the legitimate means to self-defence.

In the attacks all over Gujarat, as recorded before the Tribunal, areas were besieged for 7-8 hours, by mobs of over a few thousand (this varied in different cases but the marked similarity was the scale of the attackers). In all the cases, the leaders of the mobs co-ordinating and supervising the transport of gas cylinders, trishuls and talwars, chemicals and gelatine sticks have been identified by witnesses and survivors as prominent leaders and elected representatives from the BJP or leaders of the VHP, Bajrang Dal or the RSS. In most cases, there was large-scale mobilisation from local areas; neighbours attacked neighbours even though outsiders were called in to make up the numbers; rapes, too, were carried out by known figures from the village or locality.

This, too, was the result of definite planning, intended to terrorise completely and to destroy the faith of the survivors in co-existence or living in neighbourhoods that had been their homes, for centuries in many cases.

Women and young girls were targeted brutally, as were children. Evidence recorded before us shows how in the macabre dance of death, human beings were quartered and the killing protracted while the terrorised survivors looked on; the persons targeted were dragged or paraded naked through the neighbourhood; victims were urinated upon, before being finally cut to pieces and burnt. Hundreds of testimonies before us show how this manner and method of killing has left an indelible imprint on the minds of the survivors, who saw their near and loved ones killed and, that too, in such a fashion. These are images that have the potential to haunt, traumatise and enrage the survivors. In the case of the now well-known Gulberg society, where former MP Ahsan Jafri was killed along with 60 others (estimate of independent sources), after the housing colony was set upon, the massacre orchestrated, and the survivors had finally managed to escape in the evening, the skulls of those killed were used by some in the neighbourhood to play cricket with.

Muslim men, women and children were killed by stabbing, in private or police firing, or by burning them alive. Evidence before the Tribunal shows that the burning alive of victims was widespread. This is not accidental. For the victim community, Muslims, who bury their dead, the killing by burning was meant to annihilate as also to terrorise and establish dominance over the entire community. When 6-year-old Irfan asked for water, his assailants at Naroda Patiya made him forcibly drink kerosene, or some other inflammable liquid, before a lit match was thrown inside his gullet to make him explode within. Such brutality, which was encouraged or condoned by the government in power, is now cynically being denied.

Bodies of victims were dismembered in a merciless fashion before they were finally killed. Women and children were especially subject to this; women were not just raped but all kinds of objects and instruments were brutally inserted into their bodies. There were instances where young children, even infants, were hoisted on swords or trishuls before being flung into flames.

Unprecedented scale and degree of violence: Ethnic cleansing

The Tribunal recorded evidence from more than 16 districts of Gujarat. From the evidence placed before us it is clear that starting from February 28, within the first 72 hours [of the massacre], even as [the Chief Minister of Gujarat] Shri Modi claimed the situation to be under control, there was unprecedented loss of life and property. Thereafter, violence continued in 3-4 distinct stages right up to mid-May. Even the hearings of the Tribunal in the first half of May were preceded by warnings to call off the
Tribunal. We, too, had to ask for state security.

To cause the maximum possible damage swiftly and comprehensively, a powdery-white chemical was widely used, which not only burnt human beings to the bone, but even cement houses were completely burnt down. From Vatwa to Gulberg society in Chamanpura, Ahmedabad, to far-flung district-places like Ode, Sardarpura and parts of Vadodara, we have recorded evidence of the use of this powdery-white chemical. When Tribunal members visited Gulberg society on May 5, the compound of the society was littered with small bottles with remnants of a whitish powder inside. From Vatwa we collected not only evidence of use of this powder but also ingenious electrical wiring to ensure that all 65 homes of the Vohra Muslim Burhani society caught fire almost simultaneously. During our visit to Ankleshwar, a few days later, we recorded testimonies of many victims who said that in the attacks in that district, gelatine sticks of the kind used in mining operations were widely used.

The premeditated and meticulously planned attacks were obviously intended to ensure that the targeted homes and business establishments of the minority were reduced to bare shells. A noticeable pattern in the attacks on rural farms was the total destruction of bore-wells in such a way that it left no scope for repair of the device.

Evidence before the Tribunal shows that, guided by leaders, the trained mobs first sprinkled the targeted buildings with fuel drawn from kerbas (large cans/ barrels), or even a tanker in some cases, followed by a spray of acid. Immediately thereafter, a gas cylinder brought along by the mobs was unsealed and tossed into the flame. The result was a deadly explosion that ripped buildings apart and killed a large number of persons on the spot.

Across Gujarat, over 1100 Muslim-owned hotels, the homes of not less that 100,000 families, over 15,000 small and big business establishments, around 3000 larri gallas (handcarts), and over 5000 vehicles (private cars, trucks, taxis, autorickshaws) were badly damaged or completely destroyed in the attacks. These figures, arrived at by the Tribunal through the voluminous evidence presented before us indicate the attempt to economically cripple a community on a scale unprecedented in the post-independence history of communal violence in the country.

Looting and destruction of property 

The destruction of property across Gujarat, in the most affected cities of Ahmedabad and Vadodara, as also elsewhere, was thorough and precise. The extensive evidence before the Tribunal shows that this, too, was part of the pattern and the planning behind the attacks; to devastate and completely destroy the property of the targeted Muslim section. The Tribunal has photographs and written and oral evidence that shows how even RCC slabs of homes and shops caved in because of the intensity of the chemically-fuelled fires. As significant is the fact that every single Muslim household and business establishment was looted before being reduced to an empty shell. There are instances where, at the more affluent shops located on the main roads in Ahmedabad or Bharuch, the middle and affluent classes among Hindus, women and girls noticeably, were seen looting choice collections from a boutique or shop before it was completely destroyed. Whether it was household articles painfully collected by the working classes, or dowries that were carefully amassed over the years for girls to be married, the marauding mobs made sure that no recovery was possible and that to rebuild their lives, the affected families would, literally, have to begin from scratch.

Most of the attacks in the first round of violence began on the morning of February 28, or on March 1, the day of the Bharat bandh [public protest]. On day one of the murder and loot, brutal state wide killings were conducted with precision. Apart from that, in cities and in far-flung rural areas, evidence shows that the attacks were on the houses and business establishments of the Muslim community, which were either in Hindu dominated areas like market-places or on the outskirts of villages. This was almost like a prelude or ‘warm-up’ activity for what was to follow. In most places, the attacks started in the afternoon, driving Muslims out of their homes. From March 1-3, in all the affected villages, Muslims were forced to flee their homes taking nothing with them.

In the villages, people first tried to gather in the local mosque or in the few concrete houses that belonged to better-off Muslims. When these were also attacked, they had to flee in some available vehicles or on foot. Trees were felled to block roads and obstruct Muslims trying to escape from the frenzied, armed mob. The way in which large masses of victimised Muslims were terrorised and made to flee is despicable in a society where democracy and secularism is said to be the norm. Although in a large number of cases, people managed to escape from their villages and reach safer places, many were chased, caught, killed, and sometimes even dismembered and completely burnt. Women were stripped naked and repeatedly sexually assaulted by mobs. In many cases, the dead bodies have not been found.

Once the Muslims fled from their villages, mobs looted and then burnt their houses and shops at leisure. In many villages, houses were being torched until as late as March 10-13, and, in some instances, even later. In every structure targeted, be it a house or a shop, doors, windows, window frames, grills, electric wiring, water pipes, taps, switch-boards, electric meters, all movable property, even roofs, went missing. There were traces of the chemical powders used even when Tribunal members visited these villages two months after the crime. Every place was burnt completely. In some places, even walls have been broken down.

Elsewhere, only burnt, bare walls remained. The dwellings looked as though they had been bombed. Even bore-wells were totally damaged or blocked. Every single tree, including all fruit-bearing trees, was cut down. The marauders made sure there was no sign of life left anywhere. In most places, the looting and the destruction of property went on for days after the people had run away from their villages. Victims deposed that many of their goods can still be found in the homes of their Hindu neighbours but no attempt has been made by the state to look for them and book the culprits.

The evidence recorded before the Tribunal shows that, while Godhra provided the pretext, there was prior mobilisation of men and materials, and an organisation in place that made possible the systematic and calculated preparations that preceded many of the massacres. The mass use of gas cylinders in Ahmedabad and many other places, even while there was a shortage a fortnight before, the training needed to torch the fire-proof showroom of Harsoliya Motors (Sabarkantha), the selection of the kind of blasting devices and detonators needed to destroy Muslim-owned factories and establishments in the GIDC area in Modasa (Sabarkantha) or Vatwa (Ahmedabad), while the areas were under curfew between March 1-3; they all suggest detailed military-style pre-planning.

The Tribunal has received evidence from across the state of Gujarat that a deliberate motive behind driving Muslims out of villages where they have lived for centuries, and where an economic and social boycott is even today being carried out, is to surreptitiously and illegally take over landholdings held by them.

Military precision and planning behind attacks 

How the operations were executed: Large mobs running into thousands were led by well-known elected representatives from the BJP, leaders of the VHP, Bajrang Dal and RSS and even cabinet ministers. From the evidence before us, it is clear that these leaders quite often carried computer printouts of the names and addresses of Muslims homes and shops. Field operations were co-ordinated by a central command using mobile phones.

The formation of arson battalions: The evidence before the Tribunal clearly points to scores of key actors leading large mobs, fully aware of what they had to do and achieving their task with precision. This suggests the existence of a private, trained militia running into thousands in Gujarat. A militia, moreover, established and made fighting fit through training camps, distribution of weaponry and hate propaganda glorifying violence. Weapons used in attacks, such as swords, were of the same brand, and must obviously have been distributed in advance across large tracts of the state. The deployment in many of the attacks of large tempos or trucks, full of hired hooligans, some local and others from [Uttar Pradesh], [Madhya Pradesh] or Maharashtra, identified as such because they spoke in Hindi or Marathi, is a worrying indicator of the scale and reach of these underground operations. Village-level evidence points to hired mobs, where the hooligans were equipped with trishuls [tridents], iron rods and swords, carrying supplies of water, salted beans and peanuts and liquor pouches and paid Rs. 500 per day or Rs. 1000 per night.

Profile of the assailants: The leadership of large mobs running into thousands was provided by easily identifiable elected representatives of the BJP (including cabinet ministers), and others from the VHP, the Bajrang Dal and the RSS. From the evidence before us, it is clear that these leaders were carrying computerised sheets containing people’s names and addresses. Houses were marked off community-wise. Evidence regarding surveys collected in advance and details obtained through revenue and sales tax records, apart from electoral rolls, was placed before the Tribunal. The mobs, arriving in vehicles such as trucks, Tata Sumos, tempos, jeeps and Maruti vans, were led and directed by local Hindu leaders belonging to the Sangh Parivar [Hindu organizations]. Leaders, who used mobile phones while the attacks were being carried out, have been named by Muslim survivors in the complaints sent to the police by registered post or in the FIRs recorded.

The second rung comprised of the chief executioners who wielded all the weapons ?guns, trishuls, swords ?and handled arsenals and supplies ?petrol, diesel, kerosene, chemicals and gas cylinders ?for starting fires. They moved around in vehicles loaded with chemicals and weapons. This was the group primarily responsible for the brutal killings, sexual assaults and other abuses. Muslim survivors from many villages told the Tribunal that these aggressors carried identical backpacks filled with pouches of chemicals. The planning was so elaborate that a particular group of people had been assigned only the task of loading guns.

The third group was mainly involved in looting property from the houses and shops. In some of the tribal areas, this group consisted of Adivasis. In some villages, people said that not all of those who came in the mob spoke Gujarati. Some of them were also speaking in Marathi and Hindi.

A well-financed operation: Money, in several instances, was an added factor in mobilising mobs. The Tribunal has recorded the evidence of four witnesses who attended training camps conducted by the VHP and the [Bajrang Dal], often inside local temples.

Swords and trishuls were sold to those attending. They were indoctrinated into being prepared at all times to attack Muslims and assured that if someone lost his life performing his ‘duty’, his dependants would be paid an adequate sum of money; one witness said that a few lakhs was promised as compensation. The propaganda and the indoctrination created fanatics who were comforted by the assurance that, were something to happen to them, their family members would be well looked after.

Such access to resources raises the critical question as to who funded these operations and from where such huge resources had come from. From the evidence of expert witnesses and victims recorded before the Tribunal, it is clear that groups like the RSS, the VHP and the [Bajrang Dal] have access to large sums.

The state bandh on February 28, and the Bharat bandh on March 1 ?both called by the VHP/[ Bajrang Dal] and supported by the state BJP and the chief minister himself ?helped in the killing, loot and destruction. The fear created by aggressive sloganeering and posturing, the deathly silence and empty streets helped the trained militia to carry out their jobs with ease, unhindered by the state police.

Complicity of civil society 

With their relentless hate campaign, the masterminds of the violence ensured such complicity from civil society in their murderous deeds, that there were very few instances of members of the majority community coming out to protect Muslims. This complicity was due to the following factors.

Lack of remorse: The visible lack of remorse among a large section of the Hindu educated middle-class, about the enormous human tragedy that affected such large numbers of people in the state, is a disturbing feature of the violence in Gujarat. This situation is quite unlike that in other communal riots, where this social segment played a role in the restoration of peace. In many Hindu middle-class localities, Hindus who had social relationships with their Muslim neighbours, gave encouragement and shelter to attackers. The reality that many of these attackers were lumpen elements, of whom they would normally be fearful, did not seem to disturb them. There was enthusiastic participation of middle-class Hindus in the looting of shops. Right from the beginning of the violence, statements like, ‘a lesson needed to be taught’ and other justifications of the violence were often heard from middle-class Hindus, ranging from university teachers to petty businessmen. It is almost as if the affected people are the antagonistic ‘other’, beyond the pale of human ethics and morality. There is an eerie silence in which victims of the carnage appear to have been rendered invisible.

Fear and terror generated by threats and hate speech: The Tribunal has recorded evidence that clearly shows how Hindus who sheltered and supported affected Muslim families were threatened and abused. A witness as highly placed as Shri Piyush Desai, CMD, Wagh Bakri Chai, and a corporate leader belonging to the majority community, took the lead in organising relief and mobilising men from the trading and business groups to initiate reconciliatory measures. Even on the day he deposed before the Tribunal, May 5, Shri Desai was threatened by local VHP-[ Bajrang Dal] goons and asked to stop his activities. If a man as highly placed as him could be so threatened, imagine an ordinary citizen or a family wanting to help his/ her neighbour. Even retired High Court judges and lawyers did not have the courage to come out openly against the goons, for they, too, felt unsafe.

Tirades against peace initiatives, secularists: In their public exhortations and speeches, hate pamphlets and articles published in blatantly communal newspapers like Sandesh, and mouthpieces like Hindu Vision and Hotline, top level state functionaries in

Gujarat and their minions, have specifically targeted the small number of men and women from Gujarat and outside, who have stood out at this moment of crisis, speaking for sanity and reason, and against hatred.

Use of hate speech and hate writing 

Widespread hate propaganda was conducted through pamphlets distributed by Hindu communal organisations in different areas in large numbers. The content of these included calls for the social and economic boycott of Muslims, warnings about Muslims constituting a danger to the survival of Hindus, urging Hindus to awaken and to decimate and drive Muslims out from India.

Much of the local media played a reprehensibly partisan and inflammatory role right from February 28 onwards. Local political leaders used the electronic media in the most despicable manner. It would [also] be no exaggeration to state that the local press, particularly Sandesh and Gujarat Samachar (the former with greater impunity) was party to fuelling communal tension in the state through sensationalised, provocative, and, at times, highly inflammatory reporting.

Mobilisation of women, [tribal communities] and Dalits 

The incitement of tribal communities, and the targeting of Muslims in rural areas, is a disturbing feature of the recent violence in Gujarat. Dalits and members of the denotified tribes like Waghris and Charas were [also] active in the violence in urban areas, especially in the more gruesome instances of rape, killing and bestiality. The tragedy behind this pattern lies in the fact that influential and dominant sections of caste Hindu society have driven a wedge among the oppressed sections, pitting Dalits, Waghris and Charas against the Muslim minority. In urban Gujarat, especially Ahmedabad, Dalits and Muslims live in close proximity. The lower castes were cynically trained to indulge in violence of a kind that dehumanises the perpetrators themselves.

Women, especially from the affluent classes of Hindu society, were visible participants in the violence; in some cases, they even led the assaults and instigated Hindu men to commit sexual crimes against Muslim women.

Preparation for violence: Immediate and long term

The Tribunal recorded extensive evidence on the systematic pre-planning and preparations that also explain the military precision with which the violence was led and its devastating consequences for the state’s Muslims.

Six months before the carnage, the tempo of communal mobilisation had increased in a number of villages, with the launch of the shilapujan connected to the building of the Ram temple in Ayodhya. ‘Trishul diksha’ programmes, in which trishuls were distributed at large gatherings, were also organised in a number of areas during the same period. In almost all the affected villages, meetings were held on the evening of February 27 or on February 28 to plan the attacks.

It was only due to such organisation and pre-planning that mobs as large as 7-10,000 or more could be so quickly mobilised, not only in a large city like Ahmedabad but also in the rural areas of Gujarat.
Evidence before the Tribunal also reveals there were many cases where the Muslims fled the villages before the attacks, thanks to their being alerted in time by their peace loving Hindu neighbours. This was true especially in Bharuch, Ankleshwar and Sabarkantha districts and in parts of Panchmahal.

On the strength of the extensive evidence placed before the Tribunal, it is led to the conclusion that the Gujarat carnage has its roots in the sustained anti-Muslim mobilisation by the Sangh Parivar, among specific social groups. In the face of all the evidence of prior planning, the ‘pratikriya’ (‘ spontaneous reaction’) explanation for the post-Godhra violence touted by officials and political leaders is hopelessly inadequate, to say the least.