Construction(s) of Female Criminality: Gender, Caste and State Violence

Nikita Sonavane is a legal researcher on issues of sexual violence, indigenous self governance and anti-discrimination law and an advocate, Srujana Bej is an advocate with research interests in Aadhaar-based biometric authentication and access to socio-economic rights, and Ameya Bokil is a human rights researcher in minority rights, and criminal justice. All authors are members of the Criminal Justice and Police Accountability Project (CPA Project), a Bhopal-based research and litigation intervention.
30 August 2021

The narrative of a criminal woman finds its bearings within the caste system in India. During British colonial rule, the Criminal Tribes Act, 1871 classified several tribes as hereditary, habitual criminals who by nature were predisposed to committing petty offences. Their alleged likelihood to commit crime at any moment justified blanket surveillance against them at all times. The hereditary caste system was the primary sociological paradigm through which the colonial state understood and perceived criminality. It framed specific “deceitful” crimes as the ascribed occupations of communities that were outside the order of the caste system and pursued impure, unspecific or non-traditional occupations, sometimes without residing in permanent shelters. This article explores the gendered nature of the colonial construction of criminality attributed to women belonging to Vimukta jatis (denotified nomadic and semi-nomadic tribes) through an analysis of criminal law. The first section of this article outlines the influence of caste in the construction of women’s criminality through the CTA. The second section locates Adivasi and Vimukta women’s bodies as sites of casteist state repression through criminal law and the criminal justice system, even as custodial violence against Vimukta women by the state has been erased and made invisible. The structural underpinning of this violence is negated in “mainstream” discourse. The third section details how narratives of criminality further aid and abet repression, by analysing arrest data for excise offences in Madhya Pradesh and bail orders passed against women from the Vimukta Kuchbandhiya community in MP.

Caste and the Construction of Women’s Criminality

The British Empire’s colonial oppression of India, for capitalist accumulation of resources and wealth, was legitimised through the bluff of bringing civilisation, order and morality to the “savage” and “lawless” people of India. Tackling crime was a chief obsession of the colonial government. The limitedness of policing resources was overcome through an extensive surveillance regime, and a facade of order which pinned criminality on specific communities. These communities were easily propped up as the “proper objects of policing” on account of their socio-economic vulnerability and the “consensus” among ruling castes about their “otherness” perceived as deviance (Chandavarkar 1998). 

Furthermore, the colonial state was constantly threatened by the mobility of nomadic communities and other communities who populated the margins of “civilisation” (Rana 2011). These communities had evaded the Empire’s nefarious imposition of “modernity” and “progress” (encoded in sedenterisation, taxation and state control) through their unregulated lifestyles and economies in antithesis to the imperial capitalist enterprise. Their lifestyles, livelihoods and worlds were intimately tied to forests and common lands for hunting, procuring forest produce and trading products derived from natural raw materials. These communities’ very existence threatened the colonial state’s monopoly over forests and natural resources. To retain its monopoly and restrict communities’ mobility, the colonial state branded mobility as a criminal activity and de-historicised non-sedentary communities as dangerous people who threatened the established (industrial and social) order (Rana 2011). 

The Criminal Tribes Act (CTA), 1871 branded the “other” nomadic communities as hereditary criminals “addicted to the systematic commission of bailable offences.” The CTA’s construction of the “criminal tribe” derived its scientific veneer from Cesare Lombroso’s pseudoscientific notions of the existence of a “dangerous class” of “born” criminals who carried an innate, heritable biological trait of criminality (Bhukya 2007). Colonial anthropology in India rigorously exoticised and studied various groups. Communities which fell outside the caste order were constructed by such oppressive anthropology as descendants of an older subcontinental race with a nomadic and more savage past, following different religious cults (Bhukya 2007).  Besides, European sensibilities had long linked a nomadic life with criminality (Bhukya 2007). While Europe thus supplied ‘scientific’ rationale for criminality as a hereditary trait, the Indian caste system offered legitimacy and teleology to the branding of entire communities as hereditary criminals (Raghavaiah 1969).1 British officials believed the caste system to be the “essence” of Indian society and relied on it to justify the CTA:  

 ... people from time immemorial have been pursuing the caste system defined job-positions … So there must have been hereditary criminals also who pursued their forefather’s profession. (D’Souza 2001)

The CTA forced nomadic communities to sedentarisation in “reformatory settlements” where they were disciplined through manual labour and social alienation. Upon release, community members were required to take up permanent residence as directed by colonial authorities. All members of “criminal tribes” were subject to intrusive surveillance. They were required to register, report themselves to police authorities at fixed intervals and notify any change in or absence from residence. The police institutionalised the practice of maintaining detailed registers of these communities to document their “criminal antecedents.” CTA provisions and police regulations granted arbitrary powers to the police which were freely abused to perpetuate violence, extortion, fraud and bribery (Kapadia 1952: 102). The police were not merely colonial agents of oppression but actively served their own interests in institutionalising the ‘criminality’ of these communities by accusing them of all crimes committed in their jurisdiction, often without evidence and sometimes even in the absence of a crime. 

The archetype of the criminal Indian woman was derived from the colonial state’s zealous attempts to impose British moralities on sexuality in India. Victorian notions of female sexuality projected civilised women as “helpless, tender and frail” creatures with “every gentle feeling naturally amplified” (Radhakrishna 2001). Concomitantly, uncivilised Indian women were immoral and deceitful figures who required discipline. The very existence of any social behaviour among them which failed to conform to Victorian stereotypes of a “civilised” woman fed the colonial state’s anxieties about appearing unsuccessful in its “civilising” mission. The Infanticide Act, 1870 created the criminal Indian woman by criminalising and apportioning the entire culpability of infanticide solely on the woman, neglecting social contexts and the role of the state, male partners and the caste system (Anagol 2002). A key strategy of criminalising infanticide was recasting Indian female sexuality as being depraved, having perverted maternal instincts and no moral code (Anagol 2002). The colonial state propagated Indian women as “easily available,” “fallen from grace” and having loose or no morals (Radhakrishna 2001). 

However, the category of “women” in the colonial imagination that were considered to be ‘loose’ was further complicated by the institution of caste, wherein patriarchal relations were informed and differently constituted within graded caste inequalities (Rege 2013). Given the casteist construction of the category of Indian women, the vilification of nomadic communities as “criminal” and their extremely low status as outsiders to the traditional caste hierarchy, women from criminalised nomadic communities suffered profound negative impacts. They became particularly vulnerable to both state violence through police criminalisation and caste violence bolstered by the stigma of criminality. They were considered to have greater freedoms in sex (Kapadia 1952) (implying ‘easy’ illicit sex relations) and were characterised as being equally (if not more) unscrupulous criminals as men from criminalised communities:

women of Bedars in Bombay are often prostitutes and go as mistresses... to pave the way for burglary. The women of Harnis, similarly, often go as mistresses... and after a while disappear with the wealth of the house… The wives of Kaikadis become the temporary wives of another in the absence of their husbands who are jailed. The women of Manggarudis…cleverly pilfer clothes from the washing line, pick- pocket in the bazar and commit theft and house-breaking by day…The Waghari women are persistent beggars and cleverly steal ornaments worn by children. (MacMunn 1933; Kapadia 1952)

It is important to also note that many nomadic communities traditionally pursued professions involving performance art and entertainment activities. Given the loss of trade in the new imperialist economy and the criminalisation of mobility, some communities became compelled to depend on various other livelihoods such as agricultural labour (Bhukya 2007). Formerly, independent communities, exploited by both the revenue policies of the colonial administration and the Indian owners of capital, were forced by impoverishment to resort to some criminal activities. Their vulnerability also likely led to frequent sexual exploitation for exchange of money by the dominant and upper-caste communities, and became portrayed by colonial anthropology as a hereditary profession.  

The CTA was repealed in post-independence India, in 1952, after the Indian state deemed branding entire communities as criminal, a form of colonial barbarity. The “criminal castes” were “denotified” and claimed the self-expression, Vimukta jatis. However, various state habitual offenders acts and police regulations preserve the CTA’s spirit. These provisions grant the police wide authority to survey, search and detain “habitual offenders.” The police, an institution established by and steeped in colonial governmentality, continues to internalise the notion of inherently criminal tribes. Thus, the “hereditary criminal” of the past has only been administratively recast as the “habitual offender” of the present. The police are notoriously subservient to the caste system and patriarchy. Therefore, in cities such as Bhopal, the police often identify “habitual offenders” from Vimukta jatis. The stigma of criminality continues to be inscribed on the bodies of Vimukta communities and women, in particular, are extremely vulnerable to be persecuted through the casteist–colonial narrative of criminality. 

State violence, decreed through police surveillance and persecution, is also encoded in Brahminism when acting on the bodies of women from Vimukta and other marginalised Adivasi communities. These women are devalued and their alleged criminality is an easy excuse for the state apparatus and a casteist society to disrupt their lives, livelihoods and cultures. Their bodies are constructed as sites of criminality and deviance, perversely “entitling” the police and upper castes to wreck violence, extortion and exploitation. While upper-caste women in independent India have been able to reconfigure their identities and sexuality as being honourable due to the dictates of Brahminical patriarchy, Vimukta and Adivasi women who lie outside the caste system have not been offered the same “redemption.” The historical oppression of these communities through colonial practices and the free reign of the caste system have continued their oppression.   

Women’s Bodies as Sites of State Repression

The Brahminical legacy of a criminal justice system conceptualised by the British has been upheld through laws formulated by the post-independence Indian state. The police targeting of marginalised communities, particularly Vimuktas, has continued in a concerted manner post independence. 
This is evident from the composition of our prison population. The narratives of criminality have ensured that 66% of all prisoners are from either the Scheduled Caste (SC), Scheduled Tribe (ST) or Other Backward Class (OBC) categories and that over 18% of them are Muslims (National Crime Records Bureau 2020). The gendered dimension of the criminal justice system, in respect of how men from certain communities are targeted, is yet unexplored. The “criminality” of men from Vimukta, Dalit and Adivasi communities is constructed around stereotypes of masculinity and barbarism (Gupta 2010: 309; Hinchy 2020: 1).2 Women comprise 4% of our prison population. Scant attention has been paid to the conditions of their incarceration, their identities in terms of caste and social locations, and the narratives of criminality contributing to their incarceration.

However, Adivasi women’s bodies continue to be sites of social control in post-independence India. Inflicting violence on their bodies, particularly through sexual violence, showcases the deployment of the female body as a territorial and political battleground (Sud 2018). Rape and the disciplining of the female form are standard weapons of war in conflict areas and the regions of Bastar, Gadchiroli, etc, which are afflicted with “Maoist insurgency” or “left-wing extremism” (Bhutani Vij 2019). Dissenting women are a threat to the capitalist-patriarchy and women who dissent against ‘development’ are a double threat as they target the neo-liberal ideology. To control such women and consequently to control natural resources, the state deploys violence and creates an environment of fear in masculine logic to establish dominance over women. The sexual torture of Adivasi activist Soni Sori in police custody is an example of the chilling effect to quell dissent by Adivasi women against the Brahminical capitalist state (Women Against Sexual Violence and State Repression 2012).

The emergence of the legal category of “custodial rape’’ was following the rape committed on a 16-year-old Adivasi woman, Mathura, by police personnel in the conflict-ridden Adivasi Gadchiroli district, Maharashtra.  Mathura was branded a “shocking liar” because she was “habituated to sex” and the court concluded that she had consented to sexual intercourse with the accused policemen in Tukaram v State of Maharashtra (1979). Mathura’s case embodies how the spatial dynamics of police custody are inherently violent and unsafe for women from marginalised communities. While the mainstream analysis and protests centred around recognising this custodial rape as an act of “power rape,” neither highlighted how the violence, especially sexual violence, against Adivasi women is inherent to the criminal justice system. The discourse completely excluded Adivasi women who are in a precarious position in police presence and before the criminal justice system at large, which are both institutions of casteist violence. The cases of Mathura and Soni Sori exemplify how state violence and repression and a quest for resource control is intimately tied to the bodies of Adivasi women.

Adivasis are a heterogeneous community. Therefore, Brahminism is the point of origin of violence, but it assumes different forms in the case of Vimukta women. The construction of criminality assists the obfuscation and further perpetration of violence against Vimukta women. The police, as state agents and agents of dominant castes, routinely target Vimukta communities and sexual violence is a common tool of this targeting and oppression against Vimukta women. The state violence against Vimukta women is justified as a form of collective punishment for crimes committed by members of the community, and is in line with the Brahminical violence inflicted by members of dominant castes.

For instance, in 2007, in Betul district of MP, the murder of an upper-caste woman was pinned on persons from the Vimukta Pardhi community. In retaliation, upper-caste men and women from the village burned down the entire Pardhi community settlement. A Women against Sexual Violence and State Repression (2012: 31) fact finding uncovered that the brutal retaliation included the gang rape of 10 Vimukta women by the police . Two women from the Vimukta Pardhi community were also murdered. The use of rape in this tragedy by upper-caste men exemplifies the use of sexual violence as a tool of retribution. This tragedy is also symptomatic of the differential premium attached to different women within the caste system. Violence against upper-caste women is considered an affront to the honour of upper-caste communities which must be responded to by sexually violating women from marginalised communities. Such barbaric punishment is an attempt to control and discipline the bodies of men from marginalised communities for transgressions of the caste system.

Similarly, in Dhar district, the rape of women belonging to the Bhil Adivasi community in 2017 by the police has been buried by highlighting the “criminality” of the community, as if being criminal justifies one’s rape. Everyday policing in India is marred with violence against Vimukta women under the garb of crime control (Counterview 2017).

“Criminality” Abets Repression

The impunity of sexual and physical violence against women from Vimukta and Adivasi communities is documented even today. However, we are yet to even begin examining the quotidian procedural violence they face by the criminal justice system in the form of police surveillance, detention and registration of false cases, through the arbitrary exercise of wide discretion. Discretionary police violence is informed by practices developed by colonial government and the caste system. It extends to all members of Vimukta communities: men, women, elderly and children. Such violence further maintains the narrative of criminality and legitimises the exercise of power.

We seek to demonstrate this by analysing excise offence related policing in Madhya Pradesh (MP). We choose MP given our past criminal justice research in this region, and the MP police’s over-use of state excise laws, demonstrated by exceptionally high crime rates under the laws (Criminal Justice and Police Accountability Project 2020).3 The excise regime and its prohibition on the manufacture, sale and consumption of local indigenous alcohol such as mahua is fundamentally tied to efforts of regulating the traditional occupations of Vimukta and other communities.

We studied a sample set of 540 first information reports (FIRs) registered under the MP Excise Act, 1915 in nine police stations in three districts of MP between 2018–20 and arrest data from six districts. First, we found that excise related policing is primarily carried out against persons from certain Vimukta and Adivasi communities, though this varies across districts. Second, the percentage of women implicated in the survey of 540 FIRs was 8% (this is higher than the average percentage of women chargesheeted—5% for India and 4% for MP). Notably all but two of the 41 women implicated were from the Kuchbandhiya, Kanjar (both Vimukta) or Gond (Adivasi) communities. This suggests that of the few women who become objects of policing, only women from a few communities are disproportionately represented.

Table 1: Classification of All Women Accused Persons by Community in the Data Set of 540 Excise Act FIRs from Nine Police Stations in Three Districts of MP


Number of Women

As % of All Women

Kuchbandhiya (Vimukta)



Kanjar (Vimukta)



Total (Vimukta)



Gond (Adivasi)






Total from marginalised communities



Dominant caste






A closer examination of the above table indicated that women do not become objects of policing by accident. Some police stations targeted localities populated by a Vimukta community. This explains our finding that 25 of the 41 women were accused from one police station (Ghamapur) and nine from another (Belbagh). Thirty-three of these 34 women from Kuchbandhiya and Kanjar Vimukta communities, respectively.

Table 2: Proportion of Women Accused Persons from Five Police Stations (PS) in Three Districts of MP


Shahpur PS (Betul District)

Ratibadh PS (Bhopal District)

Betul Kotwali PS (Betul District)

Belbagh PS (Jabalpur District)

Ghamapur PS (Jabalpur District)



















Women (%)







This finding is corroborated by the analysis of all persons arrested in Ghamapur from 2018–20. Of the 1,094 excise-related arrests, 624 (57%) arrested persons were Kuchbandhiyas. Over two of every three (71%) Kuchbandhiyas arrested were women. Of the 508 women arrested by Ghamapur police in this period for excise offences, 441 (87%) were Kuchbandhiyas.  

We identify these instances as procedural violence because the excise policing regime is based on arbitrary and selective enforcement of law against marginalised community members for the offence of possession and sale of small quantities of liquor (80% of the FIRs were for quantities up to 5 litres). Furthermore, when liquor sale and manufacture licences are exclusively given to conglomerates owned by upper-caste persons, the livelihood and culture of Vimukta communities stands automatically criminalised.4 We found several instances of the same person being booked under excise offences multiple times a year; women more frequently than men. Some women have been arrested 23–28 times between 2018–20 and are now undoubtedly labelled as habitual offenders. The “habitual-offender” status entails several negative consequences including difficulty in accessing bail and increasing the likelihood of being implicated in future cases. Take for instance the 2017 rape case in Bhopal. After the case was registered, the police rounded up 20 “habitual offenders”—all men from the Pardhi community. They were released after a day’s detention when the police realised that their detention could not be justified. Therefore, the 20 Pardhi men were transferred in groups of four–five to several other police stations and were charged with the offence of belonging to a gang of thieves (Section 401 of the Indian Penal Code). The trial has been underway for over three years, and the false FIRs have been added to the list of their “criminal antecedents” further increasing suspicions of their criminality and the likelihood of future arrests (Bokil and Sonavane 2020a). Cases of thefts where the culprits are unknown are routinely pinned on men and women from Vimukta communities.

A survey of bail orders passed by the High Court of Madhya Pradesh between 2018 and 2020 in cases involving Kuchbandhiya women demonstrated that bail was denied in two of the three cases, when the woman was a “habitual offender.” In one case for anticipatory bail, the state merely claimed that 50 cases were filed against the applicant and the judge failed to examine these cases before branding the woman a ‘habitual offender’ and refusing her bail. In another third case, the “habitual offender” status led to harsher bail conditions despite the woman already spending 193 days in custody. The creation of a ‘habitual offender’ record is yet another instance of the procedural violence against marginalised communities. 

The narrative of criminality built on false stereotypes through history and in the present day has helped create police records against Vimukta women in habitual offenders’ registers. These records now breed more evidence of the ‘criminality’ of Vimukta women. 

Procedural violence may also meet sexual and physical violence against Vimukta and Adivasi women because criminal records, “criminal antecedents” legitimise the denial of liberty in courts and legitimise violence by state and upper-caste groups. An example of this is the case of Indramal Bai, a 30-year-old Pardhi woman from Bhopal who was forced into committing suicide, following weeks of police harassment. When her family approached the MP high court, after the police refused to register an FIR, the police claimed that Indramal Bai had reason to falsely accuse the police because she had “criminal antecedents” (nine cases filed against her). The court found that she had been acquitted in eight cases, save one pending case. The court directed the Central Bureau of Investigation to take over the investigation. However, the court implied that had she been convicted in even one of the cases, her death would not have deserved a fair investigation (Bokil and Sonavane 2020b). Similarly, when six Pardhi women were abducted and sexually assaulted, they were booked for alleged theft at their abductor’s shop.


Physical and sexual violence against Vimukta women is legitimised precisely by the criminal narrative constructed against the Vimukta community, itself an outcome of procedural violence. In fact, these processes are indistinct and inseparable. Vimukta communities are collectively punished by the upper castes and state alike for purported offence or harm by any one of them. This has extended to settlement burnings, mob lynchings and sexual assault. The violence Vimukta women face is legitimised by the narrative of their criminality and used to justify their extrajudicial punishment. The specter of criminality looming over Vimukta communities is a breeding ground for the Brahminical dominance of the state. The criminal justice system initiated by the police has germinated, nourished and allowed for the erasure of gendered violence against Vimukta women by sustaining the myth of their criminality. 


The authors would like to thank Bhangya Bhukya for his insightful comments on an earlier draft of the article. The authors would also like to thank Aditi Pradhan and Yashasvi Bablu Kadbe for their research support. Most importantly, the authors owe a deep sense of gratitude to the everyday resistance of Vimukta women against an oppressive criminal ‘justice’ system.

Nikita Sonavane is a legal researcher on issues of sexual violence, indigenous self governance and anti-discrimination law and an advocate, Srujana Bej is an advocate with research interests in Aadhaar-based biometric authentication and access to socio-economic rights, and Ameya Bokil is a human rights researcher in minority rights, and criminal justice. All authors are members of the Criminal Justice and Police Accountability Project (CPA Project), a Bhopal-based research and litigation intervention.
30 August 2021