Article body

The Government of India Act of 1935 was the culmination of nearly six years of legislative, administrative, and political work. The act, a scheme for a federated India of autonomous provinces and princely states, all ultimately under continued British rule, itself ran to hundreds of pages, replete with annexes, schedules, and draft legal instruments. The construction of the act required thousands of hours of testimony and deliberation in committees, parliamentary commissions, and fact-finding inquiries, as well as volumes of print: surveys, submissions of evidence, and a substantial part of recorded parliamentary debates. This proposal for Indian political reform was the signature political event between the advent of the national government in 1931 and the abdication crisis five years later. It was a consistent part of political discourse, especially as it threatened to divide the membership of Britain’s most successful political party at the time. The 1935 act’s impact was as significant as the controversy it had generated. Its provision for the election of provincial legislatures gave the Indian National Congress (INC), the party of Gandhi and Jawaharlal Nehru, a vital opening to establish firmly their authority to speak as a truly representative Indian political force. Even after 1947 and Indian independence, the act remained influential, as the new state adopted its own version of the federal framework, one that continued to feature a division of powers between the central and provincial governments with, as in the British version, significant residuary prerogatives for provincial governors and the federal centre itself.

Despite the tremendous attention this act garnered in British political life, the personalities involved in its creation and the continued importance of this legislation for the constitutional framework of contemporary India, the act has not received the sustained historical or scholarly attention one might expect. There is certainly a great deal of material to work with; in fact there may be almost too much. Incisive specialist investigations of aspects of the India Act do indeed exist.[1] The act, and the events surrounding it, appear prominently as well in larger studies of figures such as Gandhi, Churchill, Nehru, and Stanley Baldwin, and of less well-known — but significant — actors such as Samuel Hoare, Edward Irwin, and T.B. Sapru. R.J. Moore and Carl Bridge are the authors of the two very detailed examinations of the process of negotiation, calculation, and political maneuvering that led to the act. These works, exhaustive in their use of archival evidence and articulate in their delineation of a complicated narrative, stand as the best overviews of this event to date and likely will remain authoritative hereafter.[2] Yet much remains for examination, especially the question of Britain’s rationale for adopting this instrument as a way to check the growth of the nationalist movement. With that larger issue in mind, this paper focuses specifically on the working of the act between 1935 and 1937, and on British efforts to predict the impact of this plan in India.[3]

The vision many in British governing circles had of the India Act raises important questions about the ways in which the Raj worked. As those who have examined the act have noticed, its supporters held the sanguine view that the plan for federation and provincial autonomy would effectively counter a growing nationalist movement by sowing discord in the Indian National Congress and distracting Indians from the movement itself.[4] The questions, however, remain: why did these politicians and colonial officials believe it, and why did they pursue this strategy even as evidence increased against its efficacy? The answer lies in the beliefs many British élites held about India, and in the ways in which these understandings shaped, and were shaped by, methods and institutions for acquiring information in the colonial state. The mixture of colonial cultural assumptions, a sclerotic intelligence apparatus, and a dynamic Indian political environment produced a narrow, but pervasive, and persuasive, understanding of India among those most responsible for governing and sustaining the Raj. In their efforts to “make a society legible,” as the political theorist James Scott has put it, British administrators “saw like a state” and thus transformed “what was a social hieroglyph into a legible and administratively more convenient format.”[5] While Scott might assert that all states work this way, the problem of the British colonial state was not that it existed as such, but was that it was a certain sort of state, one informed by a specific culture and one underpinned by a flawed and overtaxed administrative structure. Its existence as a colonial state, moreover, imbued it with even further difficulty, notably in its problematic interaction with indigenous sources of information.

In addressing these issues, this paper engages with a larger and contested debate over colonial governance and intelligence assessment in particular, and over the creation of “colonial knowledge” in general. Christopher Bayly has addressed both concerns extensively in his Empire and Information, though it is an account which deals only briefly with the twentieth century, and he has encouraged historians to continue work on imperial intelligence, which “remains a poorly studied area.”[6] Regarding the importance of cultural belief, and its role in essentializing Indians and other colonial subjects, the events of 1935–1937 argue for the continued significance of cultural nostrums in shaping imperial policy, but also points to the failure of colonial institutions to keep pace with rapidly changing situations on the ground.

Imperial Culture and Imperial Policy

Discussions, within the Conservative Party and in India, about how best to confront Gandhi and the INC took place within an environment where there existed already a widely-disseminated and fairly comprehensive set of beliefs about India and Indians. During the period of British rule in India, notably in the late nineteenth century, there had emerged several strong ideas about what India was and what the British role there should be. These British assumptions about India were quite powerful and pervasive, so much so that they proved influential in shaping the British response to Indian nationalism. Yet these sorts of perceptions gained power because of the various limitations and flaws present in the colonial state’s effort to collect and analyze political intelligence. The Raj was hardly an all-seeing, omniscient entity, hamstrung as it was by sheer lack of personnel, inadequate training (especially in Indian languages), and an absolutely overwhelming amount of information to digest continually. Added to these was the relative isolation of many colonial civil servants and administrators, even down to the level of the police, from the vast majority of the Indian population. The small number of Indians who retained ties of one sort or another to the colonial state in fact enjoyed disproportionate influence as sources of information and guidance. In my book, Empire, Politics and the Creation of the 1935 India Act: Last Act of the Raj, I deal in detail with one such person, the Indian lawyer Tej Bahadur Sapru, whose influence might be judged by the remark of the Chief Justice of the Allahabad High Court in late 1929: “Sapru is my authority as regards Mr. Gandhi and is quite sure about it.”[7]

In such a situation cultural explanations found some hold and helped guide British thinking towards what would be the last major political reform in India before 1947: the 1935 Government of India Act, a plan for autonomous Indian provinces collected around a central government controlled in part by the viceroy and in part by an assembly of elected representatives and appointees from various princely states. The hope in New Delhi and London was that this maneuver would seriously hamper any further nationalist unity by attracting Indians worn out by Gandhian-style protests and eager for political office and patronage. The colonial assumption was that provincial differences would further hamper INC efforts to maintain a unified front, while separate delegations of Muslims and princely representatives at the centre would serve as a bulwark if the Congress should gain a significant place in the assembly. Local and provincial rivalries, the possibility of spoils of office and a lack of Indian political will to maintain a struggle for substantive goals: all of these notions rested upon well-used colonial stereotypes. The 1937 provincial elections would be the test of their efficacy.

The images of India that entered British consciousness through these various outlets shared some fundamental similarities in their presentation of Indian life, religion, and political society. It was a discourse that emphasized the idea of “difference”: not only the difference between Britain and India, but also the various differences or oppositions which existed in India itself.[8] A recurrent trope of this discourse was the difference between Western modernity and an India still plagued by “ancient” belief systems and social constructions. In short, India appeared an “authentically primitive” place.[9] The basic notion underlying many British assumptions was that the vast majority of India’s inhabitants lived in small agricultural villages; this was the “real” India, rural, placid and apolitical.[10] British administrators made something of a fetish of these rural communities. Lord Curzon saw the “real Indian people” among “the Indian poor, the Indian peasant, the patient, humble, silent millions.” They had “no politics,” but were the “bone and sinew of the country.”[11] This understanding of India found many exponents, including Richard Temple, H.H. Risley, and Katherine Mayo, who praised rural villages as the “true homes of India,” inhabited by nine-tenths of the population: “hard-working cultivators of the soil, simple, illiterate, peaceful, kindly.”[12] Indian villages were portrayed as examples of an “ancient” society, one in which change came slowly, if it came at all.[13] These portrayals placed great emphasis on the apolitical nature of these rural Indians. The interwar years also saw significant publications on this theme by Indian civil servants themselves, most notably Malcolm Darling and Frank Lugard Brayne.[14] Darling found Punjabis to be “the very marrow and soul of the peasantry,” men who had “[g]rit, skill in farming and a fine physique.”[15] The common understanding of the real India, then, envisioned a panoply of rural villages, parochial and self-contained, where life centred on the rise and fall of the agricultural calendar, and the peasantry remained blessedly ignorant of the Western modernity which had affected urban India with such unnatural results.

However, for all the praise of this peasantry’s simplicity and community, British descriptions remained largely paternalistic as well. Though inclined towards placidity or contentment, rural India was still prone to the irrationality, violence, and excitability which, in the British understanding, plagued the Eastern races. Hinduism epitomized this continued Indian backwardness. The pervasive British understanding of Hindu India contained some basic, if at times contradictory, assumptions. The religion was synonymous with weakness, both physical and moral, but was also regarded as ancient — traditional and even hidebound — and therefore a metonym for Indian society as a whole. Katherine Mayo’s Mother India, a must-read for many in British colonial circles in the 1920s — and elsewhere: it had 20 printings between 1928 and 1930 — had reinforced several stereotypes about Hinduism, most particularly the notion that it was effeminate, governed by unrestrained sexual desire, and led by “broken-nerved, low-spirited, petulant ancients” whose followers were inevitably “narrow-chested, near-sighted [and] anemic.”[16] Hindus were “small weak and timid,” according to an official source in 1909.[17] Moreover, by virtue of the caste system, Hindus remained enmeshed in a pre-modern social structure built on hierarchy and ties of kinship.[18] Even a supporter of Indian causes such as Edward Thompson (father of the historian, E.P. Thompson) admitted that caste had left India mired in “Hindu social injustice.”[19] The retired Indian administrator Rushbrook Williams summarized this view in 1938, arguing that “to many Hindus the duty owed to other members of the joint family appears something far stronger than any duty owed to the State; what Westerners call nepotism is in India a positive virtue.”[20] This immorality and lack of self-discipline also manifested itself, to British eyes, in Indians’ apparently innate aversion to honesty and preference for what Curzon called “craftiness and diplomatic wile.”[21] An overarching theme, then, was that of India’s immaturity, a condition which manifested itself in sexual license, physical and mental weakness, a penchant for graft, and an aversion to honesty and plain-speaking.[22] Moreover, this particular sense of Indian and British difference was not the only framework Britons employed to understand India, for it potently combined with another British colonial conviction, that of the numerous differences which divided India itself.

The British politician Lord Lothian, who had toured India in the early 1930s, argued that “Indian society … is essentially a congeries [of] widely separated classes, races and communities with divergencies of interests and hereditary sentiment which for ages have precluded common action or local unanimity.”[23] If British accounts sometimes presented a seemingly monolithic Hindu India, at many other times colonial assessments stressed how caste demarcations divided and fragmented this same population. This sense that caste was such a fundamental force in India pervaded British governing circles to such a degree that, for example, India Office intelligence reports on indigenous politicians began with a classification of these men’s caste status: Malaviya was a “Malwa Brahmin,” Nehru a “Kashmiri Brahmin,” and even Jinnah was described as “Mussalman (Khoja).”[24] Caste was merely one of several supposed characteristics of Hinduism, moreover, that the British used to distinguish these Indians from those of other faiths, in service of the notion that the subcontinent was permanently and historically divided by religion, and by Hinduism and Islam specifically.[25] The 1934 report of the parliamentary committee which examined the proposed Indian political reforms concluded, “Hinduism is distinguished by the phenomenon of caste … the religion of Islam on the other hand is based upon the conception of the equality of man.”[26] Indian Muslims and Sikhs garnered further respect from British officials who considered the two groups as hardier and more martial specimens, and this conviction recurred even in fictional accounts of life in the Raj.[27] Hindu weakness expressed itself not just physically, but also very much in an Indian penchant for bribery, outlandish rhetoric, and double-dealing. A popular study of India from 1934 concluded,: “The Hindu is the talker, the Mohammedan the fighter.”[28] These ideas about notable differences between Indian religious groups relied heavily not only on theological explanations, but also on theories about the impact of environment and geography on the development of divergent populations within India. These conceptions further reinforced the image of India as a land divided.

The “martial races” of northern India, Sikhs from the Punjab and Muslims from the northwest frontier, came from rugged climates that produced a hardy and forthright peasant stock, while the steamier environment of Bengal sapped its inhabitants of both moral and physical vigor, leaving them enervated and indolent.[29] Moreover, these hardier Indians came from largely rural areas, in essence, from the real India. Indian cities such as Calcutta, conversely, had produced, in British eyes, a repellent hybrid: the anglicized Indian or “babu,” an Indian, usually Hindu and not one of the martial groups, who had acquired a veneer of Western learning, but could never gain the solid moral strength of character of the Englishman. This image recurred especially in British descriptions of nationalist politicians, many of whom were Western-educated.[30] The babu, depicted as caught between his pretensions and his inalterable Indian self, was a pitiful figure, one apparently so confused about his own identity that he lived in a perpetual state of irrational excitability. It was hardly a coincidence that the figure of the babu became prominent in British imperial culture at the turn of the twentieth century, a period which saw the rise of INC nationalism in India. Whereas British commentators praised the peasants and princes who accepted their role in the hierarchy of the Raj, they castigated those Indians who wished to rise above their pre-ordained station. One British newspaper concluded that the founders of the Congress were “vapouring, gushing rhetoricians ... busy-bodies, notoriety-seekers and incendiaries.”[31]

This, then, was the image of India encountered and absorbed by those charged with meeting the challenge of a rising Indian nationalism in the years after World War I. It was a picture of India that stressed the overarching importance of difference and diversity: the gulf between a rational West and a superstitious, sentimental, and parochial East, as well as the vertical and horizontal cross-hatchings that divided Indian society against itself. This discursive understanding, combined with a particular and selective reading of Indian events, served to shape and fashion the response of British policy-makers who sought to derail any Indian attempt to undermine the Raj. This was, perhaps, the supreme test of British pretensions to know and understand Indians even better than Indians understood themselves. Would colonial knowledge save colonial rule?

Culture and Colonial Intelligence

Imperial information about India came through several channels and from multiple sources. As we have already seen, many Britons, especially in élite circles, encountered images of India and descriptions of Indian behaviour in both popular and academic discourse, as well as within certain social and cultural environments. What remains to be determined is the precise nature of the interplay among these ideas, policy-making and the colonial response to the rise of the nationalist movement. How prevalent were these ideas? Were these assumptions about India so strong in and of themselves as to overpower any other possible interpretations of Indian actions, or was their influence ultimately dictated by the intelligence-gathering abilities, or inabilities, of the colonial state?

Among those who served the Raj in India, certain beliefs about indigenous society did seem likely to exist. The bulk of this civil service came from professional families and élite educational backgrounds, with more than a few representing families that had served in India in some capacity for several generations.[32] Their educations, formal and informal, had certainly exposed these administrators to the prevailing understandings of India: its backwardness, the identification of the peasant as the truly Indian type, and the idea of an India divided many ways against itself.[33] Furthermore, service in India did still attract those who saw empire and the “uplift” of the natives, as a Christian duty.[34] Furnished with such notions, and often placed hurriedly into unfamiliar territory, many in the Indian Civil Service (ICS), unsurprisingly, appeared to frame their approach, in rural districts especially, along the lines of this received wisdom. Instead of grasping the inroads made by the INC in the 1920s and 1930s, they remained wedded to the conviction that Indian politics were only parochial, the concern really of the rural landowners more than of the placid peasantry who, without any “outside stimulus” cared “little or nothing about politics.”[35]

These ideas seem to have been especially prevalent among those serving in the ICS between the wars; they likely permeated other parts of the Indian administration, particularly the higher ranks of the Indian Police Service, as evidence adduced below should demonstrate. Brayne and Darling were well-known exponents of such a paternalistic view of the Indian peasantry and the image they helped to create, of the Punjab peasant in particular, resonated. In 1928, for example, more ICS newcomers requested a posting to the Punjab than to any other region.[36] Even if they did not gravitate to the Punjab, ICS administrators throughout rural India in these years consistently concluded that the peasantry remained apolitical and oblivious to the inducements of the nationalists.[37] As Shahid Amin has revealed, the ICS judges who presided over the trials and appeals of the defendants in the 1922 Chauri Chaura attack, where more than 20 Indian policemen died after a crowd set their station alight, refused to see the Indians’ actions as at all tied to politics and to Gandhi in particular. Instead, the defendants were “deluded peasants” who saw Gandhi as a “miracle worker”; such a vision meant the colonial state saw the event as “a common blur of rustic excesses fuelled by local political machinations.”[38] Reginald Maxwell of the ICS sensed in 1924 that the “local notables” were “nice, simple, hospitable people who are totally untouched by political agitation so long as they are understood and treated politely and feel that their position is recognized.”[39] Bolstering such conceptions was the relative distance many of these administrators experienced from Indians, whether out of racial prejudice, pressure of work, transfers, or linguistic difficulties. The report of the Simon Commission, which visited India in 1928 to examine political progress there, concluded that there was especially a growing disconnect between local District Officers and local people.[40] Henry Cotton, a prominent ICS man in Bengal in the 1880s, had recognized this much earlier, noting the instability of a system where rule lay in the hands of “a small number of foreign visitors, in a state of isolation produced by a difference in religion, ideas and manners, which cuts them off from all intimate communion with the people.”[41] An Indian who had joined the ICS after World War I remembered: “It is strange to reflect that the members of the Civil Service, whether British or Indian, hardly ever met the political leaders.”[42] As one historian of colonial rule has concluded, “the social isolation and political biases of officials often meant that the intelligence was not read aright,” and that furthermore, colonial officials spent much of their time “talking to the wrong people,” including landlords and local notables.[43]

The presence of these assumptions in colonial thinking in India may not have been, however, an indication that culture was the one truly determinative factor in official assessments of Indian behavior. Ruling India had become a massive, unwieldy, and challenging venture by the start of the 1920s. As if India’s geographical size, linguistic diversity and population were not enough, there were specific aspects of British governance that further complicated the administration of the Raj. Though the hierarchy of British rule seemed fairly simple, those who occupied these administrative positions often found themselves in great conflict, especially over issues of responsibility and authority between those in India and those in London. The Secretary of State for India, a member of the cabinet, enjoyed final authority over Indian policy, but was responsible, ultimately, not to administrators there, but to metropolitan politicians and party supporters. The India Office in London, staffed by career civil servants drawn from some of the same educational and “gentlemanly” backgrounds as the ministers they served, worked both to advise the secretary of state and to implement policy, a massive responsibility that made the India Office a huge department of state by the early twentieth century.[44] These Whitehall mandarins, such as Sir Arthur Hirtzel and Sir S. Findlater Stewart, who served as successive permanent under-secretaries between 1924 and 1937, were not members of the ICS though, their expertise residing much more in the mastery of bureaucratic processes and administration. Of course, that did not stop Hirtzel from asserting that, among other things, in India “‘the principle of contradiction’ is not a law of thought.”[45]

The politicization of the highest levels of Indian administration was another potentially disruptive, or at least complicating, factor in governing the Raj. The viceroyalty, as well as the prestigious governorships of Bombay, Madras, and Bengal, had become part of British domestic political patronage, leading to the appointment of administrators who may have been well-connected at Westminster, but who had very little experience of India. In the 1930s, for example, past and future Conservative MPs occupied all three of these governorships.[46] Such a process of appointment had serious implications, especially for the supposed independence of these influential governors from the concerns of the party leadership that had appointed them.[47] This situation also meant that, party political considerations aside, Indian policy-making at its highest levels, whether in India or in London, was in the hands of a fairly homogeneous few whose exposure to ideas about India and its inhabitants had been narrow and confined very much to specific, recurrent tropes of an élite colonial discourse that permeated the world of such Tory gentlemen.

These senior administrators did not, of course, live completely isolated lives, whether in India or in London; but their own access to information about indigenous politics and society was limited in important ways. There were several reasons why information, and political intelligence particularly, did not make its way either smoothly or accurately into the hands of imperial policy-makers. Governors, especially political appointees, and other senior administrators operated within a circumscribed world, usually meeting only Indians from the urban economic and civic élite, a system one scholar has aptly named “‘neo-darbari’ politics.”[48] (These Britons were, of course, surrounded by Indian servants and retainers constantly, but these figures hardly registered on the colonial conscience as even real persons.[49]) The ICS, upon whose presumed expertise British rule rested, was not a repository of wholly reliable information either. In some ways, this fabled “steel frame” of district officers, magistrates, and advisers seemed in danger of collapsing during the interwar years, as morale and recruitment fell, while workloads increased.[50] Already charged with immense judicial, administrative, and fiscal responsibilities, ICS men found themselves transferred frequently around India with little time to familiarize themselves with local languages, social structures, or cultures.[51] One administrator later remembered that he had learned Marathi as an ICS probationer, but was never posted anywhere that he could use it.[52]

The technical practice of intelligence collection in this colonial state remained flawed in its structure and its mechanics. As historians of the Raj know quite well, the bureaucracy of British India generated an astounding amount of paperwork: correspondence, reports, circulars, censuses and all manner of other supervisory documents.[53] Simply keeping up with the flow of information, much less analyzing and assessing it, was a Herculean task, one that distracted local officials from cultivating potential sources, and often resulted in the production of reports that were pro forma or, as Christopher Bayly has put it, “curiously ritualized documents.”[54] There were indeed efforts to construct and utilize a separate system of intelligence collection in India, building upon the institutions that had emerged at the turn of the century to counter violent nationalist activity, especially in Bengal.[55] By the end of the 1920s, there was a Central Intelligence Bureau (CIB) for the government of India in New Delhi, supported by Criminal Investigation Departments (CID) of one form or another in each province.[56] The CIB in Delhi, in coordination with British domestic intelligence and New Scotland Yard, fed information to the Indian Political Intelligence (IPI) division of the India Office in London.[57] Among the items gathered by the IPI were fortnightly summaries of local events submitted by each province, police surveillance reports, and information gathered from censoring and opening Indian mail. The system was hardly perfect though, with the Delhi CID limited to a staff of 55 in 1930, and with the simultaneous existence of three different intelligence-gathering organizations in the north-west frontier provinces until 1933.[58]

This interwar intelligence apparatus, of course, did nothing to lighten the workloads of the provincial civil servants who were expected to provide the biweekly reports which served as the nexus of the entire system. In Bombay, to take one example of what these officials faced, the secretary of the province’s home department received near-daily reports from the Bombay city police commissioner, weekly abstracts of intelligence from the provincial police generally, digests of local editorial opinion for the fortnight and, of course, accounts of revenue collection, criminal prosecutions, and other sundry affairs.[59] Limited to four or five pages, these biweekly reports could not help, therefore, but be selective, reductive, and even unintentionally misleading. In practice, these local summaries, with a few exceptions, soon came to resemble each other with some frequency, both in form and content, reporting on public meetings, local crime figures, prices, and prosecutions.[60] These reports, and others from Delhi, demonstrated furthermore a greater reliance on local newspapers and other journals than on native informants for raw intelligence.[61]

The government of India did not possess the manpower, funds, or legal sanction necessary to carry out a full-scale operation of surveillance and infiltration of the nationalist movement.[62] This was especially evident in the countryside, as in the United Provinces (UP) where, in one historian’s estimation, changes in police practice reduced the number of indigenous agents like chaukidars (watchmen) by two-thirds from 1900 to 1930.[63] The head of the provincial police admitted in 1931, therefore, that “[o]ur intelligence system has broken down and we are no longer in a position to anticipate communal disturbances and breaches of the peace in rural areas.”[64] Furthermore, those Indian police constables and officials who made up the bulk of the force were themselves limited in the information they could collect. Well-known and recognized even in cities, these Indians could not observe political gatherings unobtrusively and, as the nationalist movement grew, so did the ostracism of these men, leaving them further hampered in developing reliable sources of information.[65] The director of the Delhi Intelligence Bureau, David Petrie, recognized in some way the difficulties faced by Indian policemen, noting in 1929 the potential for “disaster if we let the Police get the idea that Government is impotent in upholding its position and authority; for we can scarcely blame any Indian for thinking that, if a Nationalist Government is about to be established, his lot will scarcely be an enviable one who has been found fighting in the last ditch for the British Raj.”[66]

The most that provincial governments, like those in the Punjab and Bombay, might do was to observe these gatherings and report that the speeches were “beyond the understanding of many” in attendance or that their participants came from the “ignorant masses and particularly youthful element [who] must be impressed by demonstrations and influenced by violent speeches delivered with impunity.”[67] The government of India was able to continue to intercept letters to and from notable Indian politicians, including both Nehrus, Gandhi, and Tej Bahadur Sapru; but this was hardly a secret to the nationalists, nor was it an effective means of judging nationalist sentiment and organization at the local or provincial levels.[68] One of Jawaharlal Nehru’s correspondents noted at the end of a letter: “I am sure the gentleman who reads this letter before it reaches you is a nice man, kind in his home etc. So I hope that, having copied the letter, he will send it on promptly to you.”[69] And, having imprisoned Gandhi in mid-1930, administrators found that no one at Yeravda Central Prison, or in the higher levels of the prison system generally, spoke Gujarati well enough to monitor any interviews the Mahatma might give.[70]

An assessment of the precise weight that cultural presumptions had on Indian intelligence gathering, both in its operation and in analyzes that flowed from it, is therefore difficult because the presence of these ideas was not the only flaw in the system. In the first place, the colonial state in the interwar years faced an overwhelming task simply in administering India, never mind trying to monitor and assess nationalist activities and the rapidly expanding Indian political press. Due to the pressure of this work, as well as some cultural presumptions about rural India in particular, district officers and others were often not in touch with representative local opinion, just as politically-appointed governors, and even viceroys, who themselves often had little experience in India, had contact with only a very limited slice of Indian society. Moreover, those departments of the colonial state dedicated to the collection and dissemination of intelligence lacked all sorts of resources to sustain an operation of such scale and scope. Thus, confronted by both a strengthening nationalist movement and a correspondingly inadequate system for monitoring or analyzing it, it was not surprising if the policy-makers of this colonial administration latched firmly onto what they understood were, and what they believed to be, the “real” ways in which India worked.

This sort of development was not without precedent in the nearly two centuries of British rule. As Bayly has noted for the nineteenth century, this “practical orientalism” often resulted from an “absence of real intelligence or a fuller understanding of the society [the British] were dealing with” and led to “officials [taking] appearances and argot to be symbolic of character and intentions.” “Information panics” grew out of the colonial state’s inability to comprehend indigenous social and cultural practices and out of the periodic rumors of insurrection that inevitably plagued a conquering colonial power, thus underlining the “limitations of colonial power and knowledge.”[71] In the 1920s and 1930s, British officialdom may not have felt panic, but neither was the Indian administration fully sure of what it had in Gandhi and the INC, nor sure how to access such information either.[72] One particular intelligence brief on Bengal strikingly articulated what this lack of reliable and consistent intelligence cost the Raj as the nationalist movement expanded:

Information is everywhere hard to come by. Police officers are insulted when they go into villages in search of it, and are refused shelter because everyone who helps a police officer is boycotted. Government is, therefore, without information of what is going on in the interior, and it would, in the circumstances at present existing in parts of Bengal, be possible for whole tracts of country so to come under Congress domination that the authority of Government would be completely ousted without the Government being much the wiser.[73]

In such an environment, officials may indeed have filled the void with presumptions or, in other words, “a series of flimsy pretexts that were always becoming texts.”[74] If poorly informed, however, colonial authorities were not always completely uninformed. What about the role of culture when officials possessed at least some concrete information? Did cultural beliefs ultimately remain the determining influence in the colonial mind, so powerful that not even masses of contradictory evidence could overwhelm them?

Congress and Provincial Elections

From 1932 through 1935, and even up to the provincial elections held as part of the reforms in early 1937, British officials in India and at the India Office stressed that some very limited concessions might go a long way in India. Again, these arguments relied on colonial conceptions of Indian, especially Hindu, character and behavior; but they also dwelled on the importance of communal and caste divisions for Britain’s long-term strategy in India. Malcolm Hailey, the dean of the Indian Civil Service, Governor of the United Provinces, and a key adviser to the India Office in this matter, saw the nationalists as inherently weak, or at least weak-kneed, and opined that many Indian politicians had realized that they had bitten off more than they could chew. He sensed among Indians that “[t]he nearer we get to [giving India] any practical measure of responsibility, the greater seems to be the hesitation to assume it.”[75] In New Delhi, Viceroy, Lord Willingdon, argued that all “the Indian” really wanted was “to be able to say before the world that he is administering his own country,” while in truth leaving the heavy lifting to British.[76] Hailey reckoned that these Indians would accept something much less than self-rule rather than take over the rigorous tasks performed by the British administration, a sentiment conveyed to the Indian Secretary, Samuel Hoare, by the Governor of Madras as well.[77] There also remained the belief that even the prospect of provincial autonomy alone in India would be enough to pull many away from the national congress movement through the prospects of offices, jobs, and patronage.[78] An India Office administrator summed up this approach to the nationalists this way: “The Bengali is an emotional creature with a bad inferiority complex …. Anything which goes to persuade him that great opportunities are open to him is all to the good.”[79] Willingdon was convinced that the lure of patronage would draw Indians into the provincial assemblies, telling a former Indian governor that “in the end [the INC] will come in — the loaves and fishes are too tempting!”[80] In the last days before the passage of the 1935 act, John Anderson, the Governor of Bengal, reported that there was “very little conviction” behind continued Indian criticism of the proposed reforms.[81] Thus informed, the Secretary of State for India found confirmation, even at the end of 1932, of his “impression that Indian public opinion … is prepared to accept a scheme provided that we keep the initiative and do not delay too long about it.”[82]

Hailey also introduced the notion that sectarian considerations, not ideas about self-rule, would dominate Indian discussion of the reforms. In India, a place where religion was supposedly paramount, “it will be the Communal Award which will be the touchstone, not questions of federation, or franchise and the like.”[83] As Hailey noted to a former ICS colleague, “the cow is still the most important figure in Indian politics!”[84] From Calcutta, the British governor reported, “what really rankles in Bengal is the Communal Award [and] the Poona Pact,” and that local politicians had told him they would work the reform scheme “for it is realized that non-cooperation will only serve to consolidate the position of the Muslims [and] Depressed Classes.”[85] The idea of a hierarchical and caste-dominated India further bolstered the arguments of those who thought it would be possible to deflect Indian political aspirations with a few well-chosen and calculated concessions. They viewed the nationalist movement as inevitably hindered by social division and stratification, with a large gulf between a small cadre of élite leaders and the ordinary Indian peasants who had been coaxed into following them. In early 1934, with Nehru imprisoned, a frustrated Willingdon noted that Gandhi continued to draw crowds, but only “because these stupid people still look upon him as a holy man.”[86] Hailey summed up the situation thus: “in the East democracy is being demanded not by the great mass of the people themselves, but by an intellectual class which hopes thereby to gain control of the people.”[87] Hailey and others predicted therefore that the more conservative Indian masses had only a weak loyalty to the Congress babus and their revolutionary ideas, and that the British might profitably deepen this split by providing some moderate political reforms.[88] For Madras, both Willingdon and the governor, Lord Erskine, saw the reforms as reinvigorating the provincial Justice Party, which would attract the votes of newly-enfranchised non-Brahmins who opposed the Brahminical INC.[89] According to these arguments, then, the proposed reforms would insure British control in India by, first, weakening nationalist resolve through the promise of office-holding and, second, by providing a political system which would encourage an Indian tendency towards localism and social division.

Between the summer of 1935 and the elections of 1937, sentiment in New Delhi and in the provinces continued to run towards the idea of eventual INC participation in the federal scheme, with the corollary that such involvement would ultimately cause major divisions in the nationalists’ ranks. In particular, a series of intercepted letters and communications fuelled official speculation that a real breach had opened between Gandhi and Nehru over the future program of the party.[90] Gandhi agreed with Nehru’s initial distaste for the new federal constitution, but the Mahatma could “not accept practically any of his methods” nor Nehru’s conviction of a coming “class war.”[91] Even if the Congress did stand in the provincial elections scheduled for early 1937, the party’s leaders had not yet decided whether they would take office and form ministries should they win. In 1936, the INC decided on the first of these issues: it would put up candidates in the provinces. Nehru, touring the country in support of this effort, found enormous crowds greeting him; “India is wide awake and expectant,” he wrote.[92] In UP, Madras, and Bombay, among other places, the INC launched extraordinary, and quite public, political mobilizations, targeting rural areas in particular.[93] The scale and scope of the Congress effort could be seen especially in Bombay, home both to urban workers and rural agriculturists. In the summer and early fall of 1936, for example, Gandhi and Congress politicians organized campaigns of political outreach in rural districts like East Khandesh, while Gandhi also inserted himself as a potential mediator into a major millworkers’ strike in Ahmedabad, drawing a crowd of 10,000 there on one occasion.[94]

When provincial governors undertook to explain the Indian political situation in 1936 to the new Viceroy, Lord Linlithgow, they continued nevertheless to underestimate Congress’s prospects and its cohesion.[95] Nearly every governor forecast that the INC might get some seats in the new assemblies, but not enough to gain control of them. Governors who had emerged from the ICS ranks, as well as those who were political appointees from London, all speculated with some confidence about political prospects in their provinces. The failures of these predictions were remarkable not only individually, but collectively as an indictment of colonial capabilities, and thus are worthy of some detailed attention. Indeed, this episode provides a real sense of the dynamic of colonial information-gathering and dissemination, and of the flaws inherent in it. It also demonstrates that, at times, even these seemingly confident officials felt the need to issue significant caveats about the analyses they provided.

Perhaps the most active of these provincial officials was, unsurprisingly, Erskine in Madras. Even before Linlithgow’s arrival, Erskine had thrown himself into an analysis of local politics. He had in the Justice Party, after all, a group he found worth watching and encouraging, and through 1935 he had Viceroy Willingdon, the former governor of the province, who remained keenly interested in its affairs. Erskine’s assessment of the situation in Madras grew more ambivalent through 1935 and 1936, likely in response to the very fluid state of provincial politics. In April 1935, he made such a case to Willingdon:

In regard to Madras, if we were to have an election this year the result would undoubtedly be a Congress majority …. It is by no means certain that there will be Congress majority in Madras when the first elections under the new Constitution come to be held. There has been a rather remarkable result in the Municipal Election at Bezwada, held 10 days ago. The Justice Party have done very well there and retain control of the Council …. But the present electorate for the Municipal Council is to all intents and purposes the same as that proposed in the Government of India Bill for the new Provincial Assemblies. So it is probable that the new voters, most of them will be non-Brahmins, will give the Justice Ministry another lease of life.[96]

Two months later, however, the governor was much more pessimistic, though his thinking would not admit any recognition of Congress’ growing strength:

The Justice Party is undoubtedly going down hill rapidly and the Congress are advancing, not on account of their own merits but simply owing to the unpopularity of their opponents. This in my view is entirely due to the supine indolence of the Ministry. The Justice Party have no organisation to speak of and the Ministers never go about the Presidency explaining and defending their own policy with the result that the Congress is having it all its own way.[97]

The prospect of an INC ministry taking office in Madras did not dismay either the governor or the viceroy, however. Both believed that Congress was eager to get back into local politics, “panting to take office” and “secure some of the spoils of office.” This meant that a Congress ministry would not act obstructively, but would do nothing “but play the game.”[98] Willingdon repeated such a view throughout 1935, in fact, judging that Congress would not be able to resist the provincial legislatures and the “desire for a share in the loot.” “[I]n the end,” he argued, “they will come in — the loaves and fishes are too tempting!”[99] A Congress victory would not mean a triumph for radicalism either, according to Erskine, for “No Brahmin with any self-respect can join the Justice Party, owing to its communal tenets, and thus the Congress in Madras is full of Brahmins who are really moderate politicians.”[100] Just before the provincial voting, Erskine’s mood swung again. He now expected the Justice Party to “maintain a firm hold on the Telugu districts.”[101] And save for “a complete landslide,” Congress would not “obtain an absolute majority over all the other parties and groups.”[102] A month later, Erskine reported, “[t]hat landslide has now taken place.” Unembarrassed by their earlier failed forecast, however, Madras authorities soon predicted that Congress would take office, and not obstruct the legislature, so as to ensure “the spate of jobs for Brahmins that their followers had been led to expect.”[103]

Hyde Gowan, Governor of the Central Provinces, saw the INC gaining only 34 out of 104 seats, largely due to the fragmentation of local politics.[104] Even when a local Indian notable warned the governor that the Congress would take over 60 seats, both Hyde Gowan and one of his district commissioners disagreed, predicting only half that number.[105] Indeed, in early 1937, local officials continued to tell the governor that the Congress was only assured of 34, though he told Linlithgow that it “passes the wit of man to guess the final result.”[106] In the end, the INC picked up 70 seats, with Gowan sputtering that the nationalists had made “absurd promises — a free house per man, halving the land revenue, and so on. A sad comment on democracy.”[107] Like Erskine, Gowan saw a bright side too. He told Linlithgow that the provincial intelligence officer had concluded that the average Congressman elected was “of very poor ability,” and that many of the “more able” men elected from the Congress were not true party devotees “at heart.”[108] The Governor of Bihar, who had argued that INC organization was more apparent than real, due to its reliance on “persons otherwise unemployable,” characterized a Congress victory there as “as much of a surprise to Congress as to their opponents.”[109]

A vivid example of these officials’ struggles with political intelligence was the case of John Hubback, presiding in Orissa. In October 1936, he had estimated that a “weaker” Congress in the region might take anywhere from 12 to 22. Two months later he was more specific, seeing 18 for the party. By January, he was hedging his guesses though, noting that if the INC took all the “doubtful seats,” it would end up with 25; but also remarking that his “District Officers … take a view somewhat over favourable to the chances of Congress candidates.” When all the votes were in, Congress had 36 seats, leaving the hapless Hubback to argue that the party was “no less surprised than others.”[110] Hubback’s problems were not, however, entirely of his own making; the information he received from the field was hardly unimpeachable, though not for the reasons he had adduced in January. For example, a district officer in Koraput in Orissa had estimated that the INC might take one of the district’s three seats; they took all three.[111]

A slightly more accurate rendering of the political scene in 1936 came from the UP, where Harry Haig had succeeded another senior ICS man, Malcolm Hailey, as governor at the end of 1934. Haig admitted that Congress popularity in the province was “definitely beyond expectation” and while he still expected non-Congress groups to get “a working majority over the Congress,” he did not believe the margin would be very great.[112] The chief intelligence analyst for the province was more forthright in admitting his inability to predict the outcome of the elections, though he did see potential in non-Congress parties:

Party politics in the U.P. are in a continuously fluid state at present …. Congress have organizational assets of great value, but the local influence of the landlords and the zamindars is still immense, and if only this influence can be fully developed to produce a strong united anti-Congress front, Congress will be hard put to it to make good its claims to popular support.[113]

The report concluded, moreover, with the type of sentiment seen elsewhere in the colonial administration. The overall picture in UP was of an “undignified scramble for seats in the new Councils, in which opportunism and personal considerations come first and all others a poor second.”[114]

Nevertheless, Haig realized the limitations of his capabilities, or at least acknowledged them indirectly. In his initial assessment for the viceroy in early 1937, Haig admitted that he was “a little out of touch at the moment,” but that recently he had had “interesting talks with Sapru and Chintamani [another moderate].”[115] Within a month of this analysis, in fact, Haig conceded that his officials might not be the best judges of the political atmosphere. Having met and listened to a non-Congress landholder and office-seeker, the governor warned Linlithgow: “I am inclined to think that candidates are at the moment in closer touch with what is going on in the villages and certainly in the minds of the villagers than our district officers are, and I feel that things may have gone rather further than we officially realize.”[116]

One of the few officials to match Haig in obliquely noting his own limitations was Michael Keane in Assam, who reported in January 1937:

I have scarcely seen any of the three Indian Members of my Government for the last month or more. They are all away in their constituencies and have too much on their minds to bother about files. Administration is in fact being conducted in the old bureaucratic style and no one seems any the worse for it.[117]

The overall result was INC victories in many provinces and, ultimately, the ministries in seven of 11 provinces under coordinated party control. It was, Zetland wrote, “a much greater measure of success for the Congress than any of us anticipated.”[118] Nevertheless, both in London and in India, officials put on a brave face and argued that, all appearances to the contrary, the election did not signify a true popular ratification of the Congress. This sort of rationalization had appeared even before all the votes were cast. Herbert Emerson, a long-serving ICS man and governor of the Punjab, saw “a great stirring of the political consciousness of the masses,” but also doubted whether any effects of this mobilization would “be more than temporary.”[119] James Sifton, governor of Bihar, offered the “broad view” that the results were “evidence of the weakness of their opponents rather than of the strength of the Congress party.”[120] The viceroy, however, offered the boldest re-reading of the evidence: “It is easy, however, to exaggerate the permanent effect of such a campaign and, notwithstanding the success of the Congress in the elections of certain Provinces, I am inclined to think that they are still far from having such an organization and such a unity of purpose as would make them as formidable as they would like to be.”[121]

Post-election analyses returned to the assumptions that had led administrators to think that Indian Hindus, in particular, were incapable of such a feat. The fault lay with the “greatly enlarged and very ignorant electorate,” one gripped by “extreme sentimentality and timidity,” and one that had fallen for all of the Congress’ “absurd promises.”[122] Sifton believed that 90 percent of the voters there saw the ballot box simply as a “letter box for Gandhi.”[123] Apart from these condemnations of the Indian electorate, a few other themes emerged in officials’ post-mortems. There was some grudging acknowledgment of the Congress’ organization, but this was often couched in the complaint that other parties had not prepared themselves at all for the elections.[124] Brabourne lamented a lack of “party discipline” in a non-Brahmin rival to the INC, while Haig criticized UP-landed élites who had fallen victim both to arrogance and to “petty personal feuds.”[125] Zetland was more positive when he told the cabinet that “among parties and groups other than the Congress the real work of party formation is still to come.”[126] The Indian secretary may have found encouragement for such a sanguine judgment in the viceroy’s own rather optimistic sense that the election results did not show deep anti-British feeling: “It is [Linlithgow wrote] only to the extent that the notion of taxes is linked to ‘Government’ that there has been any direct anti-Government (and therefore, to some extent, anti-British) prejudice raised in the villages.”[127]

An even more common sentiment was that the Congress victory was more problematic for the party than a defeat might have been, especially as the Congress leadership had not yet decided on accepting office. Some in the party seemed in fact quite eager to do so: they would “not indefinitely allow themselves to be deprived of the power and the privileges to which they have been looking forward.”[128] In the viceroy’s estimation, this left Nehru and Gandhi “both concerned to prevent provincial autonomy breaking up the all-India unity and discipline of Congress.”[129] Linlithgow still believed “in the potency of Provincial Autonomy to destroy the effectiveness of Congress as an All-India instrument of revolution.”[130] The “provincial outlook” of many Indian politicians would in the end fracture the present “artificial unity in the Congress ranks.”[131] Furthermore, colonial officials contended that such electoral success had shocked many in the INC, at both the national and provincial levels, leaving the “more intelligent amongst them … to regret the wildness and multiplicity of the promises to which they are committed!”[132] Hyde Gowan had “a feeling that Congress is slightly awed by the situation” and “the serious responsibilities of Government.”[133] Zetland’s officials at the India Office concurred, arguing that Congress had wanted only enough votes to become an “effective nuisance without caring for any responsibility.”[134] Now the INC had to deliver, but in an environment in which its support, according to Haig, was “very wide but not yet deep.”[135] A few months after the election, only a few British officials seemed to have noted the larger lesson learned: “There are fissiparous tendencies in the Congress, which superficial observers are inclined to think must soon break the movement, but past history shows that any such idea is an illusion.”[136] The viceroy, once characterized by Nehru as “heavy of body [and] slow of mind,” remained sure that office-holding would bring a more moderate Congress, interpreting the actions of the nationalists as a sign that they had “so entirely and unconditionally accepted our point of view.”[137]

All of these explanations for Congress victories, as well as many of the optimistic assessments of the pitfalls such success presented the nationalists, found their way from India to London, and via Zetland, to the cabinet itself. The illegitimacy of the vote, the false promises made by Congress to electors, and the potential split in the party over taking office, all of these informed the memoranda Zetland prepared for his colleagues. As an explanation of Congress success, he offered a picture of rural India provided by Linlithgow: “The villager, trained by circumstance to respect power, voted for the party who appeared at the moment to possess power.”[138] As an assessment of the future, the secretary argued, “The dilemma is one for the Congress rather than for the Government and there is no occasion for us to help them out of their difficulties.”[139] And as a vision of Indian opinion, Zetland presented to the cabinet the image of “the vast host of silent Indians sitting on the fence waiting to see which way the cat was going to jump.”[140] Among the cabinet, Hoare was one of the few to offer any advice or reaction to Zetland’s report, urging the secretary and the viceroy “to maintain contact with the Congress in the Provinces and to seize any opportunity that offered of persuading them to take office at any time.”[141] Hoare still seemed to believe in the efficacy of his scheme.

The 1937 elections illuminated many aspects of the British approach to political intelligence in India. The analyses offered by governors, viceroys, secretaries, and denizens of the India Office showed the great confidence — at times a confidence undiminished by previous misreadings — many of these officials displayed in discussing Indian politics and in making sweeping and speculative pronouncements. Especially well-assured were those conclusions and statements that drew heavily on assumptions of Indian behaviour, old imperial nostrums that possessed a real resilience in the official mind. These assessments drew on other sources too, of course, but here again they demonstrated some of the limits of colonial information-gathering. The India Office and the cabinet relied on officials who were actually in India, but who utilized fairly limited resources. The viceroy and his governors had some access to Indian opinion, but only of an élite sort and, apparently, only intermittently. Even when these superior officials did receive what turned out to be sound advice, either from Indians or from lower-ranking officers and collectors, they were as likely to discount or even disregard it as to pass it on. Complicating the process of intelligence collection even further, of course, was the fact that events in India, especially when subject to a superficial or partial reading, provided just enough support for what were often ill-judged or ill-informed analyses.

There certainly existed tension within the INC leadership over the question of office acceptance, and over the distinction between taking office to govern and taking it so as to “wreck” and obstruct and further colonial reform schemes.[142] Nehru was himself well aware that his embrace of the latter strategy, as well as his strong socialist sympathies, put him at odds with many of his colleagues.[143] Eventually, as Sapru observed, Nehru “yielded to the strength of public opinion among his followers” and accepted the creation of Congress ministries in the provinces, though Nehru also remained on guard for what he termed “counter-revolutionary” tendencies among these politicians.[144] Those in the INC who wished to accept office, and to govern thereafter, had a variety of reasons for doing so. Failure to take office might allow other, smaller parties to “consolidate their power and make [the INC’s] work more difficult.”[145] In particular, Congress politicians in Madras held this view quite strongly. There, as Hermann Kulke and Dieter Rothermund have noted, the party had not yet consolidated the substantial gains it had made over the past five years against the Justice Party, and office acceptance likely seemed an attractive way to solidify the Congress position.[146] To the notion that politicians were eager to enter legislatures solely for their own benefit, those in Madras responded that the Congress membership had already agreed to cap ministers’ salaries, meaning that it was “not the desire for emoluments of office but honest conviction that thereby we can advance the struggle for Swaraj” that drove such a policy.[147] Despite the potential for splits in the Congress in 1937, therefore, these fissures did not actually occur. Through Nehru’s flexibility and willingness to compromise, Gandhi’s efforts at intra-party diplomacy and, perhaps most important, the enthusiasm of provincial Congress parties for governing as a committed, idealistic party, Congress was able to take office and work coherently at both the local and national levels.[148] A national “High Command” coordinated the work of provincial governments, and in turn these local politicians continued to build grass roots support for the party.[149] The 1935 act had not, as its proponents hoped, divided the party; in fact, it had done the opposite. After 1937, the INC could claim ever-increasing national support, as it built on its electoral victories and used political power to attract and bind the population to it. Thanks to the India Act, the Congress could also claim something more valuable than organization, though. In an election held on a larger franchise than ever before — 30 million men and women (15 percent of the population) were eligible to vote and nearly half did — they had done well in both urban and rural areas. The party now had significant popular validation, complicating one of the Raj’s most consistent defenses against the nationalists: the argument that the INC did not represent India. The expansion of the electorate had not, in the end, weakened Congress, but had given the party an opportunity to display its actual strength and to suggest what it might do in a fully democratic India.