Hostname: page-component-6766d58669-l4t7p Total loading time: 0 Render date: 2026-05-17T21:16:01.832Z Has data issue: false hasContentIssue false
  • English
  • Français

Caught In-Between: Coerced Intermediaries in the Jails of Colonial India

Part of: Special Issue: Punishing Workers, Managing Labour

Published online by Cambridge University Press:  22 February 2023

Michaela Dimmers Open the ORCID record for Michaela Dimmers [Opens in a new window]
Michaela Dimmers*
Affiliation:
Max Weber Forum for South Asian Studies, Delhi, India; Centre for Modern Indian Studies, Modern Indian History, Georg-August University, Göttingen, Germany, e-mail: michaela.dimmers@cemis.uni-goettingen.de
Rights & Permissions [Opens in a new window]

Abstract

This article analyses the role coerced intermediaries had on colonial power and authority in the prisons of British India. Coerced intermediaries in this context were convicts placed in positions of control by the colonial prison administration as warders, overseers, and night watchmen and night watchwomen, summarized here under the term “convict officers”. These convict officers were employed by the colonial authorities to maintain a coercive order and became essential to the exercise of colonial authority and control in the prisons of British India. The article argues that with their employment, the colonial administration created a third group within its prisons, situated between the colonial administration and the inmates. This contradictory practice blurred the lines of colonial control and authority and raises larger questions about intermediation by unfree and coerced people in unfree and coerced colonial contexts. The focus here is not so much on what intermediation is but on what it does. At the same time, the article relates the system of convict officers as intermediaries to the theoretical concepts used by Foucault and Goffman and questions the binarity used in most of their theories.

Information

Type
Research Article
Information
International Review of Social History , Volume 68 , Special Issue S31: Punishing Workers, Managing Labour , April 2023 , pp. 237 - 251
Check for updates
Copyright
Copyright © The Author(s), 2023. Published by Cambridge University Press on behalf of Internationaal Instituut voor Sociale Geschiedenis

This article studies the exercise of colonial power and authority as well as its reliance – partially but substantially – on intermediate groups. It does so by observing the system of convict officers, examining how the colonial state used convicts and how, at the same time, the system could be utilized by convicts for their own purposes within the colonial prison system.Footnote 1 The question asked here is not so much what intermediation is but rather what it does. Moreover, it examines how the colonial state exercised its authority in an institution emblematic of colonial coercion. To maintain this authority, it is argued, the colonial regime engaged in a contradictory practice of employing convicts, the very people it convicted of crimes and incarcerated, in order to control the prison population and maintain a coercive order, thus blurring the lines of colonial control and authority.Footnote 2 This system was not specific to a colonial context. A system of coerced people watching and exerting control over other coerced people existed in diverse oppressive regimes, such as slavery,Footnote 3 and contexts of incarceration, such as concentration camps, gulags, labour camps, or penal settlements. In the nineteenth-century London Newgate prison, the post of wardsman existed, an individual “chosen by the keeper or by the prisoners” to maintain “a rough sort of order”, distribute bread and water, and settle disputes among prisoners.Footnote 4 This system was disallowed in 1865Footnote 5 but was available as a model for the employment of convict officers in colonial India. In Vietnamese colonial prisons in the nineteenth and twentieth centuries, convict officers were called contremaîtres or caplans.Footnote 6 Nazi concentration camps employed so-called Funktionshäftlinge or Kapos; that is, Kameradschaftspolizei, or comradeship police. Kapos had different positions within the German concentration camps, including heads of labour squads and camp institutions, as well as foremen.Footnote 7 These are just a few examples. In each context, slaves and inmates employed in such or similar positions constituted an intermediate layer between coercers and coerced. As such, they can be defined not only as coerced intermediaries, but also as coerced coercers. The existence of such groups also questions the binarity in the theoretical concepts used by Foucault and Goffman on total or disciplinary institutions.

Taking the point of departure of studying the convict officer system raises issues not only of intermediation in unfree and coerced colonial contexts but also of labour, outlining the limits of coercive systems relying on intermediation and the possibilities of carving out a certain level of freedom for coerced labourers in such contexts.

Coerced Intermediation

The convict officer system was introduced in mainland India in the North-Western Provinces and was, according to the report of the Jail Conference of Inspectors-Generals of 1877, fully developed there in British India.Footnote 8 Even though the reasons given in the governmental sources for introducing and retaining such a system vary somewhat, it is primarily budgetary justifications that are specified throughout.Footnote 9 The first textual trace of convicts holding supervisory powers over their fellow inmates in these provinces dates from a report published in 1861, which describes a system of punishment for prison offences that was put in place in Agra Central Jail and developed into a shift system for sentry duties by prisoners in their barracks.Footnote 10 From its inception, the convict officer system was expanded to other central jails in the province and adapted to the colonial regime's changing needs in carceral punishment and labour. Rules were laid out, marking the beginning of a regulated convict officer system, at least on paper. With the exception of the regulation of timings for meals and head counts, these rules appear to have been almost the same as for the staff guards, regulating their uniform, privileges, salary, selection, and duties.Footnote 11 Within a few years, this system became an established part of the prison complex, so much so that by 1867, the Superintendent of Benares Central Jail stated that he was looking “forward to the time when the Jail will be entirely guarded by prisoners; for there can be no doubt that only under such a system will the Jail discipline ever become perfect”.Footnote 12 By 1888, convict officers guarded work gangs inside and outside the prisons, in workshops and factories, to such a degree that by then, “[t]he whole watch and ward by day and night, the overseeing of the factories and workshops, and the control of the out-door gangs is practically carried out by the convict warders, superintended by the free warders”.Footnote 13 This practice was confirmed by Sardar Sahib Bhai Ganda Singh, jailor of Benares Central Jail, during an interview with members of the Indian Jails Committee 1919–1920, in which he stated that “[a]ll the gates and all the factories are in the hands of convict officers”. Ganda Singh answered in the affirmative when asked if “the jail is practically run by convict officers”.Footnote 14 In 1896, the duties for three grades of convict officers were described as follows: (1) the convict night watchman, considered the lowest rank, who worked “as an ordinary prisoner during the day and keeps watch for two hours during the night in the barrack”; (2) convict overseers, responsible for labour crews or gangs, who worked during daytime; and (3) convict warders, who took “the place of a paid warder in case of a vacancy”. Only one convict warder for every eight or nine staff warders was supposed to be employed. The number of convict overseers was officially limited to “5 per cent. of the jail population”. Convict night watchmen, however, could be employed in an unlimited number and on an ad hoc basis.Footnote 15 A female convict warder had to be employed in every central and district jail, but her duties were not defined in the jail manuals. Like her male counterparts, she received a salary and so-called remission marks. She was supposed to be given accommodation outside the female barracks for the night and call for – presumably male – assistance when needed.Footnote 16 The duties of all ranks of convict officers were adapted throughout the years.

As they occupied a crucial role within the colonial jails of British India, their service became an intrinsic and essential element of the penal system on which the colonial prison administration became reliant in guarding prisoners, overseeing labour, enforcing discipline, and maintaining control. Their employment constituted a specific part of the broader prison labour system. Their power over fellow inmates was comparable with that of prison officers recruited from the extramural population. Convicts employed in this unfree labour system occupied a privileged position but remained imprisoned inside the system, situated between staff members and prisoners, occupying an ambivalent position in the microcosm that was the colonial prison system. This ambivalent position could create a rupture within the prison population, as convicts in such positions had to invariably adjust their loyalties towards their fellow prisoners and the prison administration, de facto and de jure their employer. At first glance, the convict officer's position could be seen as that of an intermediary agent or intermediary, comparable or equal to those working in such positions in free(er) contexts, who, as numerous studies of intermediaries in colonial India have shown, were employed in a multitude of areas and came from all strata of the colonized society.Footnote 17 David Arnold portrays the system of convict warders as “a replication within the prison order of a factor common in Indian labour organization – the use of foremen, known by such terms as sardars and maistries, to supervise the work of others”.Footnote 18 However, it can be argued that there were intrinsic differences, for example in the power of negotiation, between intermediaries such as sardars or maistries (and other intermediaries for that matter) and convict warders, or convict officers as such, for the simple fact that free intermediaries could change employers or end their contract, where applicable. Divergence in the disciplining power merits consideration as well. Free(er) intermediaries could more often than not determine the level and space of intermediation they could occupy within societal hierarchies, which most convict officers could not do due to the coercive space they were situated in. Considering the coercive and unfree contexts they were detained in, they cannot be put on a par with intermediaries in free(er) circumstances per se, even if some of the functions of their role were similar to, or even the same as, those of free(er) intermediaries. The unfree and coerced context in which they worked made their position intrinsically different from intermediaries outside the prison. This suggests an overarching question: how can people in an unfree context, here incarceration, be considered intermediaries within or between the system of oppression and the oppressed? The same could be asked, for example, about enslavement and indentured labour.Footnote 19

Ambiguous Alliances

There were at least three structural levels of inmate control in the colonial prison that comprised primarily physical and psychological control: (1) disciplinary control: guards and rules, constituting the administration's disciplining intentions, which could have been officially or unofficially enforced;Footnote 20 (2) spatial control: architectural features, such as walls, bars, doors, and gates that could be locked and unlocked at the administration's will; and (3) taxonomical control: bureaucratic procedures and forms of administrative control, manifested by registration and documentation of the prison population, permeating a prisoner's personality and history, being part of a wider taxonomical project by the colonial regime.Footnote 21

Disciplinary control was an essential part of power since it wielded authority over people's movements, labour, food, speech, sleep, hygiene, and health, encompassing all aspects of everyday life. Convict officers had an evident power over their fellow prisoners since they played a central role in the everyday mechanisms of inmate management and constituted a large part of the control mechanisms of everyday life exercised in the name of the colonial state. They were placed – or perhaps caught – between the prison administration and the prison population. As shown below, when illicit behaviour became known, they held complex and discretionary forms of power over the prisoners as they could decide what information to pass on to the prison administration and what to withhold, at times being caught between the two groups, as the notion of “hopelessness”, cited below, partly exemplifies.Footnote 22 Being part of the prison population, they had an intrinsic insight not only into the prisoners’ intramural lives but also into mechanisms of resistance against the prison administration. This gave them some control over the prison administration as informants and, possibly as mediators between the inmates and the prison administration. As they were part of both the administration and the inmate population, they could or had to choose to whom to be loyal: the former, the latter, or themselves. They had to or could constantly determine with whom to collaborate and with whom not, be it during escapes, riots, and disturbances, or when it came to other breaches of jail discipline. This power, be it to inform on other prisoners, prevent or aid escape attempts, or even to decide who was allowed to get up at night, exemplifies the control these convicts could exert and epitomizes the ambivalences in loyalty convict officers could have towards both fellow prisoners and the administration. Evidence about these convicts’ relationship or standing with other (free) staff and other convicts is scarce in the sources. In one report Inspector-General of Prisons Clements alluded to a tense relationship between prisoners and convict officials:

Control of the convicts is mainly vested in convict officials. Some of the convicts in our jails are not amenable to discipline and resent being ordered about by fellow convicts and they give trouble if the shoe pinches too tight. This was a direct factor in the genesis of two disturbances in the central prisons, one at Benares and the other at Lucknow, and recently there was a third riot in the Allahabad central prison. Some convicts received serious injuries and as a result four died.Footnote 23

Unsurprisingly, the available sources remain generally opaque concerning the question of why such breakdowns in jail discipline occurred. Was it due to general dissatisfaction with prison conditions or policies that boiled over, singular changes in policies, or certain events that led to protests and clashes between prisoners and staff, including convict officers?Footnote 24 The account just cited suggests that assaults had taken place, at least on this occasion, due to a resentment of the power and privileges convict officers had. While there is no a priori reason to presuppose an overriding sense of solidarity between prisoners, nor that they would have worked together as a unified group against the prison administration, it can be assumed that power relations, hierarchy, class, and caste played a large, if not dominating, role in the way prisoners, confined in a limited space, lived together. David Arnold has suggested that some of these power relations at play outside the prison walls might have differed inside the prison, as some convict officers considered low caste could have held positions of power over higher caste inmates; in other words, this system may have inverted caste hierarchies.Footnote 25

Alliances appear to have been constantly forged, renegotiated, and broken. At times, the convict officers were on the side of the prison administration, as is illustrated by examples that include an incident in which a convict officer killed a fellow convict to protect a head jailor,Footnote 26 as well as instances of extrajudicial ill-treatment of other prisoners or the prevention of attempted escapes. At other times, they were on the side of the prisoners, especially when facilitating escapes, or not reporting infractions of jail rules. MacTaggart, Inspector-General of Prisons, stated that even though convict officers could “generally be trusted to prevent serious breaches of discipline such as assaults and escapes as far as they can” and were always “loyal to jail superintendents in emergencies”, they would “not report minor breaches of discipline”, would “do little or nothing to prevent prisoners having forbidden articles”, and could not “reasonably be expected to do so”.Footnote 27 Consequently, their use of discretionary powers to maintain jail discipline, on the one hand, and balancing any kind of allegiance between their fellow prisoners and the prison administration, on the other, seem to have been tolerated by at least parts of the prison administration. At the same time, rules in the jail manuals prescribed that special remission was to be given to any convict for collaborating with or aiding the prison administration in certain ways, for example, by “protecting an officer of the prison from attack”. This reveals that breakdowns in prison discipline, perceived or real, were anticipated.Footnote 28 One obvious form of collaboration with the prison administration concerned the organized ill-treatment of prisoners. Allegations that the higher echelons of the prison administration were witnesses to or ordered convict officers to perpetrate extrajudicial ill-treatment were voiced in the interviews of the Jail Committee Report of 1919–1920.Footnote 29 In a later report, the United Provinces Jail Inquiry Committee of 1929 reported having looked into such allegations. The committee stressed that no evidence of alleged ill-treatments, such as beating inmates with a shoe or putting them into stress positions, could be found. Even though it admitted that “unauthorized punishments” would occur, the committee believed this was not common or severe in the provinces.Footnote 30 Sachindra Nath Sanyal, incarcerated in connection with the Benares Conspiracy and Kakori Train Dacoity cases,Footnote 31 reported to the committee instances of ill-treatment by convict officials, but not by other officials. Convict officials, he is quoted in the report, would “pick up quarrels with the prisoners over whom they want to exercise overlordship” and beat them. Allegations by other convicts were voiced as well.Footnote 32 Jawaharlal Nehru, who was to become the first prime minister of India after independence and was incarcerated multiple times during the independence struggle, also described the ill-treatment of a prisoner who had allegedly bitten an assistant jailer and who was severely beaten presumably by convict warders and guards, whom he called “pukkas, lambardars &c.”, on the order of the assistant jailer. He further asserted that convict officers would beat convicts “just to please the jailer”.Footnote 33 But Nehru also described instances of solidarity among convict officers with other prisoners; for example, when during a hunger strike by inmates that also involved a labour strike in Naini Prison, convict officials asserted that even though they would perform their usual or official tasks, they would not mistreat fellow inmates if ordered.Footnote 34 These examples exemplify not only the violence prevalent in colonial prisons and breakdowns in prison discipline, but also the fine line of loyalty between the administration and prisoners that convict officials apparently tried to tread. This fine line put the convict officers, especially higher ranking ones, at an intersection of the regulated and the regulator.Footnote 35

Disciplining Loyalties

Privileges and punishment were not only used to ensure loyalty; both were also indicators of carceral power structures and hierarchical differences between convict officers and inmates within the colonial prisons, using, in effect, the divide-and-rule approach that was so characteristic of colonial authority. Control over the labouring inmate body was thus ensured by incentivized labour through privileges and mechanisms of punishment. The prison administration evidently felt that privileges were important to maintain “a high standard of discipline and encouraging good work” as the “safe custody of prisoners” would depend “to a great extent on the reliability and steadfastness of these officials”.Footnote 36 Privileges for convict officials included a salary, an official tobacco allowance, or being allowed to purchase and smoke tobacco, wearing shoes, being paraded and sleeping separately from other inmates, not wearing fetters, and being allowed to grow a beard and their hair “to such extent as the Superintendent may think sufficient as a mark of distinction from the ordinary Convicts”, having better or different food and cooking for oneself, as well as keeping pets.Footnote 37 Their special status was also reflected in the frequency of the heavily regulated meetings with the outside world. By 1902 prisoners could, as per the jail manual, have only one to two interviews per year; convict officers could have one per month.Footnote 38 However, the most important privilege seemed to have been the increased good conduct or remission marks they received, which meant they could gain certain privileges and an earlier release. Other convicts received good conduct marks and some of these privileges as well, but generally, convict officials received them to a much greater extent.Footnote 39 The contraposition of privileges was punishments, which could include revoking privileges and “warning, forfeiture of good conduct remission, suspension, fine and dismissal”,Footnote 40 putting the prison administration into an even more powerful position as the possibility of punishment was presumably ever looming. Such punishment within punishment constituted an effective tool of control not only for convict officers, but for all labouring convicts. As the ARCMJ reports show and David Arnold has established, staff officers committed infractions, just as convict officers did.Footnote 41 Staff warders could be punished with demotion, fines, dismissal, and imprisonment, and were, as the statistics provided in the ARCMJ reports suggest, punished on a higher level than convict officers.Footnote 42 This accentuates the notion that convict officers were more reliable than staff warders, who were considered “mostly men of inferior class and in many instances untrustworthy”.Footnote 43 However, the language of reliability, trustworthiness, and loyalty used in the government reports changed gradually. If convict officers were described in the ARCMJ report of 1863 as “superior in trustworthiness and efficiency”, they were by 1901 believed to be “fairly trustworthy” and carried out their duties “very fairly satisfactorily”.Footnote 44 By 1907, their behaviour was considered still “on the whole very good”, but at the same time, it was held that they

undoubtedly very frequently connive at or, wilfully shut their eyes to the introduction of prohibited articles, especially tobacco, but it is doubtful if in this respect they are much worse than many of the paid warders. The convict overseer no doubt generally discovers the hopelessness of his position in opposing himself against the united determination of his fellow prisoners to obtain such articles.Footnote 45

The passage illustrates, on the one hand, the compromises the prison administration was willing to make to keep this system in place and, on the other, that convict officers were not mere instruments of the jail administration but navigated the prison system and the various pressures with which they were confronted. They did so also in order to tap into the structures of privilege and incentives offered by the administration for compliance with its needs. In turn, the colonial state maintained these structures of privilege and incentives by these same needs, in effect granting convict officers an extraordinary form of power in part due to their ambivalent intermediate position. Even though a lack of sources echoing their voices directly makes it difficult to analyse in detail how they negotiated this nexus of power and powerlessness, it is apparent that they could use the trust placed in them to their advantage.Footnote 46 Preventing escapes could lead to an immediate release, a reward that was given, for example, to one convict warder who, by holding the inner prison gates closed, had prevented an escape attempt involving a convict overseer. Two other convict warders, who came to his help, received “two months special remission”.Footnote 47 At the same time, if an infraction against prison rules was detected or even suspected, the punishment could be swift and severe. Considering that the post of convict officer elevated its holder to the legal status of a public servant, they were punished not as convicts but as employees of the colonial state if they broke rules in their capacity as convict officers. Most infractions for which details were documented in the ARCMJ reports were related to escapes, for example, the case of a convict warder and a staff warder who were both suspected of aiding two convicts to escape and were sentenced to “four years’ imprisonment”.Footnote 48 In another case, a convict warder escaped with a convict, aided by a “convict overseer on duty outside”, who was sentenced for his involvement to “one year's simple imprisonment and Rs. 100 fine or three months more simple imprisonment”. Other officials apparently involved in the case were also punished.Footnote 49 More rarely documented infractions committed by convict officers included the case of two convict officers who were transferred from Benares to “Agra Jail for insubordination”, one of whom had apparently tried to kill a European officer and was sentenced to “ten years additional imprisonment”.Footnote 50

With this system, the colonial state created an elite group of convicts within its prisons that was expected to be overall loyal to the state and prison administration. As shown, incentives, mainly in the form of privileges, were offered to retain this loyalty in order to establish and maintain a system of carceral power structures exerting control and punishment over the prison population that was cost-effective and could affirm, either in reality or as a conjecture, the reasoning of convict reform. With it, the colonial regime had constructed a powerful mechanism for coercion, control, and jail discipline in its prisons and, considering the extent of incarceration, for a large part of the colonized population.Footnote 51 Both the prison administration and the convict officers benefited from this system. The benefits to the colonial regime seem to have been much greater since the convict officer system gave it the resources to control parts of the colonized population, thus if not enabled then at least helped it to expand, develop, consolidate, and maintain its control over colonial India.Footnote 52 Convict officers were used to ensure parts of this imperial order. Their employment helped the colonial regime establish a high level of control over the prisoners. Yet, by putting convict officers in effect in charge of jail discipline, the colonial prison could serve as an illustration of how the British empire was, indeed, “an ultimately precarious shifting and unruly formation, quite distinct from its self-projected image as an orderly behemoth”.Footnote 53

Conceptual Observations

The very existence of such an intermediate group complicates Foucault's and Goffman's theories. It adds to the already established critique of their work that suggests a more complex understanding of how power was exercised, for example, in institutions such as the colonial prison.Footnote 54 Both Foucault and Goffman based their analyses of institutions such as prisons on a binary structure comprising the administration/staff and the inmates and did not consider any third group, such as convict officers.Footnote 55 Goffman even speaks of a “staff-inmate split” as “one major implication of the central features of total institutions”.Footnote 56

When it comes to an intermediate inmate group, constituted of coerced coercers, punished labourers, and punishing labouring inmates, the axiom of the resulting binary theorem of prison organization as total institutions (Goffman) or of institutions disciplinaires (Foucault) must be developed and extended.Footnote 57 Foucault's approaches to, for example, docile bodies and heterotopia also fail to take into account such intermediary groups and thus the complexities of actual sociopolitical systems, in this case, colonial prisons. His vision of the docile body, which can be “subjected, used, transformed and improved”,Footnote 58 seems somewhat doubtful in light of the examples about convict officers. Since the colonial regime's approach to convict officers aimed to exploit their labour as a means to economize budgets, those convict officers were certainly “subjected” and “used” too. It cannot be denied that there was also, at least officially, an intention to “reform”. The question here is how disciplined and docile people put into the position of discipliner (and at the same time constituting a disciplined entity) were on a pendulum constantly swaying between disciplining and being disciplined, subject and object of the prison administration, hence of control and power? At the same time, the Foucauldian analysis of power, especially his critique of the binary formation of power, helps to understand these complex and constantly adapted power relations. In his discussion on “power and strategies”, Foucault shows that “one should not assume a massive and primal condition of domination, a merely binary structure with ‘dominators’ on one side and ‘dominated’ on the other”, but rather “a multiform production of relations of domination”.Footnote 59 Convict officers were part of this “multiform production of relations in domination” as they were simultaneously the dominators and the dominated. The dependency of the colonial state on convicts to ensure the intramural security of their prisons and enforce discipline among the inmates implies a more nuanced picture of how power and control worked, relying in part on intermediation with specific features arising from its unfree settings, which, in turn, were partially shaped by the colonial context. It shows that absolute control was not always in the hands of the colonial state but perforated with different actors who had – at times – differing interests.

While it might seem self-evident that convict officers were intermediaries, it is not self-evident whether their functional role was that of agents of colonial or penal intermediation, or both.Footnote 60 As mentioned, convicts employed in this manner were considered public servants by law. They were not mere passive subordinates but active participants in this system, constituting an elevated position within the hierarchy and using their position for their own purposes to gain privileges and freedoms in an otherwise unfree and coercive environment.Footnote 61 As the example of convict officers demonstrates, rather than being a static process, this intermediation could be a vastly complicated and complex maze of shifting loyalties, privileges, and negotiations.

Conclusion

Convict officers negotiated the seemingly contradictory state of having great power and being simultaneously powerless; they disciplined yet were disciplined, were controlled yet were agents of control over their fellow prisoners. They constituted an in-between group within the prison hierarchy, considered their benefits and used the system to their advantage, but also negotiated a pendulum of loyalty to the prison administration and their fellow prisoners. Even though only institutional records could be used here, and the voices of those who constituted the larger part of convict officers are rarely heard, it becomes clear that at least some without a voice outside the prison walls might have had some power inside them. The unique position they had can be defined as an intermediary coerced body that fulfilled some functions of intermediation between the colonial prison administration and their fellow inmates, though without possessing the full range of characteristics associated with colonial intermediaries in free(er) contexts – underlining the differences between various forms of intermediation. What is dealt with here comprised different registers of intermediation, between which a fine balance had to be struck by convict officers to avoid punishment by the prison administration and hostility from fellow prisoners. The colonial prison administration had to tolerate sporadic breakdowns in its authority to ensure the broader need for prison discipline and, by way of punishment, could, for the most part, hold this breakdown of authority in check.

Overlooked in Foucault and Goffman's concepts on total or disciplinary institutions, such an in-between group challenges and calls for a loosening of their binary approach. It complicates their theories and suggests a more complex understanding of how power was exercised in an institution such as the colonial prison.Footnote 62 Adding the layer of coerced coercers challenges their binary theorems, but combining it with the Foucauldian analysis of power and its “multiform production of relations in domination” shows the complexity of power relations in the everyday and the omnipresent need for compromise in coerced contexts, be it in the colonial prison, penal settlements, or other coercive institutions. In essence, the logics of power had to be constantly assessed, not only by the coerced coercers, but also by the coercers themselves. Looking at how these additional layers of domination worked in such contexts not only sheds light on the logics of such coercive institutions; it also shows that it could lead to significant and often unforeseen or even unforeseeable ramifications on the processes of exercising this authority and its impacts.

Moreover, registers of penal intermediation and colonial intermediation were, at times, at odds. What intermediation precisely does is rarely captured in its conceptual understanding, which tends to highlight aspects of collaboration and divide-and-rule techniques of exercising authority and often neglects to inquire into its actual outcomes. A more nuanced approach to the question of carceral labour, punishment, intermediation, and the term intermediary itself, seems necessary. Convict officers were undoubtedly able to exert power over prisoners and were thus part of the executive force of the colonial punishment regime, but were at the same time subjugated to the same regime. The convict officer system in North Indian colonial jails thus exemplifies the need to look into the manifold forms of intermediation and its complex and nuanced results, especially when it is connected to questions of punishment and labour. The issue of intermediaries in unfree and coercive circumstances within South Asia and beyond needs further research to identify the finer textures of not only such intermediation but also modes of power and control in order to better understand the complex structures of labour and punishment in such contexts.

Footnotes

I am grateful to Ravi Ahuja and Christian G. De Vito for their comments and insights. Adam Fagbore, Douglas E. Haynes, Indivar Kamtekar, Nabhojeet Sen, Samita Sen, the participants of the Modern Indian History Research Seminar at the Centre for Modern Indian Studies, University of Göttingen, and of the interdisciplinary conference “Punishment, Labour, and the Legitimation of Power”, organized by the Bonn Center for Dependency and Slavery Studies, University of Bonn, 18–19 February 2021, as well as the reviewers, added greatly with their comments and perspectives. Sebastian Schwecke endured endless discussions and read numerous draft versions. All comments gave form and shape to this article. This work would not have been possible without the generous support of the Deutsche Forschungsgemeinschaft (DFG).

References

1 “Convict officers” for the purpose of this article includes convict night watchmen and watchwomen, convict overseers, and convict warders. The sources sometimes also include convict teachers, convict clerks, etc. in the term convict officers. There was no distinction made between jails and prisons in colonial India. This article will limit itself to the prisons of colonial India in the North-Western Provinces, later United Provinces.

2 For criticism of this system, see Government of India, “Report of the Indian Jail Conference assembled in Calcutta in January–March 1877, under the Orders of His Excellency the Governor General in Council: With Appendices”, IOR/V/27/170/2 (IOR = India Office Records) (Calcutta, 1877), pp. 70–72 (for the whole discussion on convict officers, see pp. 69–75); Arnold, David, “The Colonial Prison: Power, Knowledge and Penology in Nineteenth-Century India”, in idem and David Hardiman (eds), Subaltern Studies VIII: Essays in Honour of Ranajit Guha, 6th edn (New Delhi, 2008), pp. 148–187, 154Google Scholar; for more criticism, see e.g. Government of India, “Report of the United Provinces Jails Inquiry Committee”, IOR/V/26/170/5 (Allahabad, 1929), pp. 105–108. Sofsky has argued that the employment of inmates in concentration camps was “blurring the line of distinction between personnel and prisoners”; see Sofsky, Wolfgang, The Order of Terror: The Concentration Camp, 2nd edn (Princeton, NJ, 1999), p. 98Google Scholar.

3 Wiethoff, William E., “Enslaved Africans’ Rivalry with White Overseers in Plantation Culture”, Journal of Black Studies, 36:3 (2006), pp. 429–455CrossRefGoogle Scholar; Walker, Timothy, “Slaves or Soldiers? African Conscripts in Portuguese India, 1857–1860”, in Chatterjee, Indrani and Eaton, Richard Maxwell (eds), Slavery & South Asian History (Bloomington, IN, 2007), pp. 235–261Google Scholar.

4 Ignatieff, Michael, A Just Measure of Pain: The Penitentiary in the Industrial Revolution, 1750–1850 (London, 1978), pp. 39–40CrossRefGoogle Scholar.

5 United Provinces Jails Inquiry Committee Report, 1929, p. 114.

6 Zinoman, Peter, The Colonial Bastille: A History of Imprisonment in Vietnam, 1862–1940 (Berkeley, CA, 2001), esp. ch. 4Google Scholar.

7 Niethammer, Lutz, Der gesäuberte Antifaschismus. Die SED und die roten Kapos von Buchenwald. Dokumente (Berlin, 1994)Google Scholar. A Funktionshäftling was an “operational inmate” or an inmate with certain functions within the concentration camps of Nazi Germany. Sofsky, Order of Terror.

8 Report of the Indian Jail Conference, January–March 1877, p. 69; even though the sources show clearly that the system was introduced in the North-Western Provinces into mainland colonial India, the report of the Indian Jails Committee of 1919–1920 states otherwise: Government of India or East India, “Report of the Indian Jails Committee 1919–1920 (Contents: Sir A.G. Cardew chairman. Vol. I Report)”, I (Report and Appendices) (London, 1921), p. 68; for earlier convict systems outside the Indian mainland, see e.g. Anderson, Clare, Legible Bodies: Race, Criminality and Colonialism in South Asia (Oxford, 2004), pp. 29–30Google Scholar; Arnold, David, “Labouring for the Raj: Convict Work Regimes in Colonial India, 1836–1939”, in de Vito, Christian G. and Lichtenstein, Alex (eds), Global Convict Labour (Leiden, 2015), pp. 199–221, 217Google Scholar; McNair, John Frederick Adolphus, Prisoners Their Own Warders: A Record of the Convict Prison at Singapore in the Straits Settlements Established 1825, Together with a Cursory History of the Convict Establishments at Bencoolen, Penang and Malacca from the Year 1797 (Westminster, 1899)Google Scholar.

9 See e.g. Government of Bengal, “Proceedings of the Third Conference of Inspector-General of Prisons held at Calcutta, 1927”, IOR/L/PJ/6/1961, File 2014 (Calcutta, 1928), p. 16.

10 S. Clarke [presumably Inspector-General of Prisons Stewart Clark, mentioned in the next footnote], “Prison Returns of the North Western Provinces, for 1860 [1859]”, IOR/V/24/2029 (Allahabad 1861), p. 8. In the following, these yearly reports will be abbreviated as ARCMJ (Annual Report, with tabular statements, for the year x on Condition and Management of the Jails in the North-Western Provinces (later the United Provinces)), as these reports were titled from the report for the year 1861 onwards. Shelf mark, archive, and page numbers will follow in that order.

11 ARCMJ, 1863, IOR/V/24/2030, British Library, pp. 18–21; Report of the Indian Jail Conference, January–March 1877, pp. 69–70. Even though the jail manual of 1863 for the province did not mention convict officials, the jail administration report for the same year introduced rules for convicts employed as officers. See Clark, Stewart, A Manual of Jail Discipline and Economy for the Use of Officers in Charge of Jails in the North-Western Provinces (Agra, 1863)Google Scholar.

12 ARCMJ, 1867, IOR/V/24/2031, British Library, Appendix No. III: Reports of the Superintendents of Central Prisons in the North-Western Provinces, Annual Report of the Benares Central Prison for the year 1867, Hooper, Superintendent, p. 71. Not everyone within the echelons of the prison administration was in favour of the system. However, dissenting voices were rare. For criticism of the system from within the prison administration, see Report of the Indian Jail Conference, January–March 1877, pp. 70–72. In this connection, see Arnold, “Colonial Prison”, p. 154. For more criticism, see United Provinces Jails Inquiry Committee Report, 1929, pp. 105–108. The members of the Indian Jails Committee 1919–1920 were divided in opinion concerning the system: the voiced general criticism can be found in Indian Jails Committee 1919–1920 Report, 1921 (Cardew Vol. I Report), pp. 68–69.

13 ARCMJ, 1888, IOR/V/24/2036, British Library, p. 8.

14 Government of India, “Indian Jails Committee, 1919–20, Volume V, Minutes of Evidence taken in the United Provinces, The Punjab and the North-West-Frontier Province, Calcutta, Superintendent Government Printing, India”, IOR/L/PARL/2/407 B, 1922, p. 1305.

15 ARCMJ, 1896, IOR/V/24/2037, British Library, p. 12. The Jail Committee of 1889 had also complained that different terms were used in different provinces and found that there was a variation in grades: W. Walker and Alfred Swaine Lethbridge, “Jail Administration in India (Walker and Lethbridge) Committee 1889: Report”, IOR/V/26/170/2 (Calcutta, 1889), p. 50.

16 Major Charles MacTaggart, “Rules for the Management and Discipline of Prisoners in the N.-W. Provinces and Oudh”, IOR/V/27/171/91 (Allahabad, 1902), p. 278.

17 See e.g. Ravi Ahuja, “Networks of Subordination: Networks of the Subordinated: The Ordered Spaces of South Asian Maritime Labour in an Age of Imperialism (c. 1890–1947)”, in Ashwini Tambe and Harald Fischer-Tiné (eds), The Limits of British Colonial Control in South Asia: Spaces of Disorder in the Indian Ocean Region (London, 2009), pp. 13–48; Bates, Crispin and Carter, Marina, “Sirdars as Intermediaries in Nineteenth-Century Indian Ocean Indentured Labour Migration”, Modern Asian Studies, 51:2 (2017), pp. 462–484CrossRefGoogle Scholar; Rana P. Behal and Prabhu P. Mohapatra, “‘Tea and Money versus Human Life’: The Rise and Fall of the Indenture System in the Assam Tea Plantations 1840–1908”, Journal of Peasant Studies, 19:3–4 (1992), pp. 142–172; Rajnaryan Chandavarkar, “War on the Shopfloor”, in Rana P. Behal and Marcel van der Linden (eds), India's Labouring Poor: Historical Studies, 1600–2000 (Cambridge, 2012), pp. 265–278; Eugene F. Irschick, Dialogue and History: Constructing South India, 1795–1895 (Berkeley, CA and London, 1994); Amit Kumar Mishra, “Sardars, Kanganies and Maistries: Intermediaries in the Indian Labour Diaspora During the Colonial Period”, in Sigrid Wadauer, Thomas Buchner, and Philip R. Hoffmann-Rehnitz (eds), The History of Labour Intermediation: Institutions and Finding Employment in the Nineteenth and Early Twentieth Centuries (New York, 2015), pp. 368–387; Norbert Peabody, “Cents, Sense, Census: Human Inventories in Late Precolonial and Early Colonial India”, Comparative Studies in Society and History, 43:4 (2001), pp. 819–850; idem, “Knowledge Formation in Colonial India”, in Douglas M. Peers and Nandini Gooptu (eds), India and the British Empire (Oxford, 2012), pp. 75–99; Tirthankar Roy, “Sardars, Jobbers, Kanganies: The Labour Contractor and Indian Economic History”, Modern Asian Studies, 42:5 (2008), pp. 971–998; Sebastian Schwecke, “Merchants, Moneylenders, Karkhanedars, and the Emergence of the Informal Sector”, in Harald Fischer-Tiné and Maria Framke (eds), Routledge Handbook of Colonial South Asia (Routledge, 2021), pp. 145–155; Samita Sen, Women and Labour in Late Colonial India: The Bengal Jute Industry (Cambridge, 1999); idem, “Making Coolies: Labour Brokerage and the Tea Industry in India, 1830–1930”, digital presentation given to the German Historical Institute London, 28 April 2020; idem, “Commercial Recruiting and Informal Intermediation: Debate over the Sardari System in Assam Tea Plantations, 1860–1900”, Modern Asian Studies, 44:1 (2010), pp. 3–28; Subramanian, Lakshmi, “Banias and the British: The Role of Indigenous Credit in the Process of Imperial Expansion in Western India in the Second Half of the Eighteenth Century”, Modern Asian Studies, 21:3 (1987), pp. 473–510CrossRefGoogle Scholar. Just to name a few.

18 Arnold, “Labouring for the Raj”, p. 217.

19 I am grateful to Christian G. De Vito, Nabhojeet Sen, and Samita Sen for the discussion and their insights on the topic.

20 David Arnold considers colonial medicine as “an agency of disciplinary control”. Arnold, “Colonial Prison, p. 180.

21 On taxonomy, see e.g. Anderson, Legible Bodies; Arjun Appadurai, “Number in the Colonial Imagination”, in Carol A. Breckenridge and Peter van der Veer (eds), Orientalism and the Postcolonial Predicament: Perspectives on South Asia (Philadelphia, PA, 1993), pp. 314–339; Nicholas B. Dirks, Castes of Mind: Colonialism and the Making of Modern India (Delhi, 2004); Peabody, “Cents, Sense, Census”; idem, “Knowledge Formation in Colonial India”; Radhika Singha, “Settle, Mobilze, Verify: Identification Practices in Colonial India”, Studies in History, 16:2 (2000), pp. 151–198; Smith, Richard Saumarez, “Rule-By-Records and Rule-By-Reports: Complementary Aspects of the British Imperial Rule of Law”, Contributions to Indian Sociology, 19:1 (1985), pp. 153–176CrossRefGoogle Scholar.

22 For the notion of the enabler and the traitor, see David Turnbull, “Boundary-Crossings, Cultural Encounters and Knowledge Spaces in Early Australia”, in Simon Schaffer, Lissa Roberts, Kapil Raj, and James Delbourgo (eds), The Brokered World: Go-Betweens and Global Intelligence, 1770–1820 (Sagamore Beach, MA, 2009), pp. 387–428.

23 ARCMJ, 1927, IOR/V/24/2040, British Library, p. 21.

24 For examples of breakdowns of jail discipline due to changes in the prison policies, see Anderson, Clare, The Indian Uprising of 1857–8: Prisons, Prisoners and Rebellion (London, 2007), pp. 37–54CrossRefGoogle Scholar; David Arnold, Colonizing the Body: State Medicine and Epidemic Disease in Nineteenth-Century India (Berkeley, CA, 1993), pp. 110–111; idem, “India: the Prisoners’ Revolt”, IIAS Newsletter, 39 (2005); idem, “India: The Contested Prison”, in Frank Dikötter and Ian Brown (eds), Cultures of Confinement: A History of the Prison in Africa, Asia and Latin America (London, 2007), pp. 147–184, 163–165; idem, “Colonial Prison”, pp. 150–152, 171–172; Rachna Singh, “Messing, Caste and Resistance: The Production of ‘Jail-Scapes’ and Penal Regimes in the Early 1840s”, in William Andrew Pettigrew and Mahesh Gopalan (eds), The East India Company, 1600–1857: Essays on Anglo-Indian Connection (Abingdon/New York, 2019), pp. 193–217; Singha, Radhika, “‘No Needless Pains or Unintended Pleasures’: Penal Reform in the Colony, 1825–1845”, Studies in History, 11 (1995), pp. 29–76CrossRefGoogle Scholar; idem, A Despotism of Law: Crime and Justice in Early Colonial India (Delhi, 1998), pp. 278–284; Yang, Anand A., “Disciplining Natives: Prisons and Prisoners in Early Nineteenth Century India”, South Asia: Journal of South Asian Studies, 10:2 (1987), pp. 28–45CrossRefGoogle Scholar; idem, “Of Lotahs and Men: Confronting of the Body (Politics in the Lotah Emeutes of 1855”, in James H. Mills and Satadru Sen (eds), Confronting the Body: The Politics of Physicality in Colonial and Post-Colonial India (London, 2004), pp. 102–117.

25 Arnold, “India: The Contested Prison”, p. 172. This could also be the case with intermediaries such as the kangani, as Crispin Bates and Marina Carter point out, referring to Peebles: see Bates and Carter, “Sirdars as Intermediaries”, p. 465.

26 Government of India, House of Commons, “Death of an Indian Prisoner Detained in Benares Central Jail: Result of a Post-mortem; Parliamentary Enquiries”, IOR/L/PJ/6/1946, File 2745, 1927.

27 ARCMJ, 1910, IOR/V/24/2039, British Library, p. 9. For another example, see ARCMJ, 1881, IOR/V/24/2035, British Library, p. 8.

28 MacTaggart, “Rules for the Management”, p. 266; Government of the United Provinces of Agra and Oudh, “United Provinces Jail Manual, Revised Edition 1927”, IOR/V/27/171/93 (Allahabad, 1927), p. 37; Government of the United Provinces, “United Provinces Jail Manual: Containing the Rules for the Superintendence and Management of Jails in the United Provinces, Revised Edition 1941”, IOR/V/27/171/95 (Allahabad, 1942), p. 57. The jail manuals for the years 1927 and 1941 prescribe that in this and other cases special remission was to be given to those who were eligible for “ordinary” remission and with some exceptions to those not being eligible for remission.

29 Indian Jails Committee 1919–1920 Report, 1922 (Vol. V Minutes of Evidence), p. 1324. Not everyone who gave evidence for this report confirmed this (see below).

30 United Provinces Jails Inquiry Committee Report, 1929, pp. 158–162, 281–285. However, the chairman of the Jail Inquiry Committee and the Committee believed there would be “some” extra-legal ill-treatment, as found in the registers for staff, who were punished for beating prisoners, but prisoners in interviews in the absence of jail officials had apparently not alleged ill-treatment. The chairman of the committee stated that he believed prisoners would not consider a slap in the face or beatings with a stick or cane as “very serious”, or worth a complaint. He considered such treatment as “most wrong and most improper” and did not condone any “infliction of unauthorised punishments by members of the staff”. Not all committee members shared his opinion. According to another member, any superintendent of a jail who would not detect severe ill-treatment would “fail his duty”. See United Provinces Jails Inquiry Committee Report, 1929, pp. 159–161, 284–285.

31 On the Benares Conspiracy case, see e.g. Government of India, “Benares Conspiracy Cases: Judgments of the Special Tribunal”, IOR/L/PJ/6/1434, File 1381, 1915–1918; On the Kakori case, see Government of India, “The Kakori Train Dacoity Case: High Court Proceedings and Judgments; Newspaper Extracts”, IOR/L/PJ/6/1910, File 2226, 1925–1927.

32 United Provinces Jails Inquiry Committee Report, 1929, p. 280. For descriptions of other attacks, see pp. 166–169.

33 Jawaharlal Nehru, Selected Works of Jawaharlal Nehru (New Delhi, 1973), IV, p. 342.

34 Ibid., p. 366.

35 For the concept of “the regulated and the regulator”, see Joshi, Chitra, “Public Works and the Question of Unfree Labour”, in Stanziani, Alessandro (ed.), Labour, Coercion, and Economic Growth in Eurasia, 17th–20th Centuries (Leiden, 2013), pp. 273–287Google Scholar.

36 ARCMJ, 1925, IOR/V/24/2040, British Library, p. 10.

37 ARCMJ, 1863, IOR/V/24/2030, British Library, p. 20; ARCMJ, 1889, IOR/V/24/2036, British Library, p. 8; ARCMJ, 1895, IOR/V/24/2037, British Library, p. 10; MacTaggart, “Rules for the Management”, p. 278; United Provinces Jail Manual (Rev. Edn 1927), pp. 46–47; United Provinces Jails Inquiry Committee Report, 1929, p. 67; United Provinces Jail Manual (Rev. Edn 1941), p. 68. The ARCMJ and the Jail Committee's recommendations give an insight into the application of these rules in the everyday life. See also Arnold, “India: The Contested Prison”, p. 172. On pets, see Jawaharlal Nehru, An Autobiography, 6th ed. (Delhi, 1988), p. 358.

38 MacTaggart, “Rules for the Management”, pp. 217–219. The rules mention only interviews with friends and do not specify if for interviews with families the same rules would apply. See also Indian Jails Committee 1919–1920 Report, 1922 (Vol. V Minutes of Evidence), p. 1290.

39 On remission marks, see ARCMJ, 1889, IOR/V/24/2036, British Library, p. 8; convict warders could earn double the remission marks of an “ordinary” convict. For statistics on prisoners who received an additional thirty good conduct marks for “good service as Lumberdars”, and those prisoners who received fifty additional good conduct marks for “good service as Warders”, see ARCMJ, 1869, IOR/V/24/2032, British Library, pp. 68A–69A.

40 ARCMJ, 1925, IOR/V/24/2040, British Library, p. 61A, as just one example. The ARCMJ also include documentation and statistics for the punishment of all prisoners and staff, see also footnote 42.

41 Arnold, “India: The Contested Prison”, pp. 171–172; Arnold, “Colonial Prison”, pp. 154–155.

42 See ARCMJ for the years 1887, p. 25; 1888, pp. 8, 26; 1889, pp. 8, 29; 1890, pp. 8, 21 (all IOR/V/24/2036, British Library); 1891, pp. 17, 39; 1892, pp. 23, 50; 1893, pp. 9, 31; 1894, pp. 10, 46; 1895, pp. 11, 52; 1896, pp. 12, 33; 1897, pp. 20, 83; 1898, pp. 13, 36 (all IOR/V/24/2037, British Library); 1899, pp. 14, 36; 1900, pp. 13, 39 (both IOR/V/24/2038, British Library).

43 ARCMJ, 1898, IOR/V/24/2037, British Library, p. 10.

44 ARCMJ, 1863, IOR/V/24/2030, British Library, p. 12D (Simson, Secretary to the Government of the North-Western Provinces, No. 1876 A. of 1864, dated 27/06/1864); cf. ARCMJ, 1901, IOR/V/24/2030, British Library, p. 4.

45 ARCMJ, 1907, IOR/V/24/2038, British Library, p. 7.

46 For example ARCMJ, 1891, IOR/V/24/2037, British Library, p. 16.

47 ARCMJ, 1921, IOR/V/24/2039, British Library, p. 7.

48 ARCMJ, 1912, IOR/V/24/2039, British Library, p. 6.

49 ARCMJ, 1922, IOR/V/24/2040, British Library, p. 8.

50 ARCMJ, 1864, IOR/V/24/2030, British Library, p. 50c.

51 See also Yang, “Disciplining Natives”, pp. 30–31.

52 See also ibid., p. 30.

53 Ashwini Tambe and Harald Fischer-Tiné, “Introduction”, in idem, The Limits of British Colonial Control in South Asia: Spaces of Disorder in the Indian Ocean Region (London, 2009), pp. 1–10, 3. For the colonial prison as a visualization of colonial power, see e.g. Clare Anderson and David Arnold, “Envisioning the Colonial Prison”, in Dikötter and Brown (eds), Cultures of Confinement, pp. 304–331; Arnold, “India: The Contested Prison”; Waits, Mira Rai, “Imperial Vision, Colonial Prisons”, Journal of the Society of Architectural Historians, 77:2 (2018), pp. 146–167Google Scholar; Singha, Despotism of Law.

54 See also Peabody, “Knowledge Formation in Colonial India”, p. 98, on the need of expansion of the binary categories used in the wider colonial context. I am grateful to Douglas E. Haynes for his comments and insights (not only) in this regard.

55 For the colonial Indian prison and the critique of Foucault see the work of Clare Anderson, David Arnold, Radhika Singha, and others.

56 Goffman, Erving, “Characteristics of Total Institutions”, in Walter Reed Army Medical Center, National Research Council (eds), Symposium on Preventive and Social Psychiatry (Washington DC, 1958), pp. 43–93, 47Google Scholar; Goffman, Erving, Asylums: Essays on the Social Situation of Mental Patients and Other Inmates (Harmondsworth, 1961), p. 9Google Scholar.

57 Goffman, Asylums; Foucault, Michel, Surveiller et punir. Naissance de la prison (Paris, 1975)Google Scholar. In these works, Foucault and Goffman apply a fundamental binary character (i.e. inmates and staff) of total/disciplinary institutions.

58 Foucault, Surveiller et punir, p. 138 (English translation by Alan Sheridan, Discipline and Punish: The Birth of the Prison (New York, 1979), p. 136). I am grateful to Indivar Kamtekar for his comments on this and other topics.

59 Quoted in Gordon, Colin (ed.), Power/Knowledge: Selected Interviews and Other Writings 1972–77 – Michael Foucault (New York, 1980), p. 142Google Scholar. See page 62 for a discussion on “agents of liaison”.

60 This notion is derived from Corsín Jiménez in the completely different context of trust. Jiménez, Alberto Corsín, “Trust in Anthropology”, Anthropological Theory, 11:2 (2011), pp. 177–196, 178–179CrossRefGoogle Scholar.

61 MacTaggart, “Rules for the Management”; United Provinces Jail Manual (Rev. Edn 1927); United Provinces Jail Manual (Rev. Edn 1941). For last three sources, see Section 23 of the Prisons Act, Act IX of 1894, in the respective jail manuals.

62 I am grateful to Douglas E. Haynes for his comments and insights here.