I

It is helpful to begin at the point where the comparison was made most forcefully: with the three major comparative works which were all published between 1900 and 1914. At first glance, these essays all highlight Rome's imperial achievement and recommend it as an instructive example for the British to learn from in their government of India. Sir Charles Lucas, for example, describes Rome as ‘incomparably the greatest Empire of the ancient world’,Footnote ¹⁵ and Lord Cromer echoes his judgement, wondering at the diverse ways in which the Romans ‘left their own abiding mark on the destinies of mankind’.Footnote ¹⁶ At the start of his essay, Bryce too dwells at some length on Rome's unparalleled ability to win the loyalty of its subjects and ‘make out of the mass of races and kingdoms that had existed before the Macedonian conquest, a single people who were at once a Nation and a World Nation’.Footnote ¹⁷

However, in terms of how the Roman example might be instructive for Britain in the government of its own empire, all three essays develop a nuanced approach. While all outline intriguing similarities between the two dominions in terms of size, population, and the direct manner of government, and claim that of all other possible comparisons Rome is the most apt, all three writers are equally aware of some quite fundamental differences between conditions in the Roman empire and British India.

In particular, they point to the much more serious barriers of colour and religion which the British faced in their Indian empire.Footnote ¹⁸ For the most part, there was no colour bar in the Roman empire. Although there was substantial contact with peoples of a different colour (most notably black Africans), the Romans, they argue, possessed no philosophy which discriminated against people on this basis.Footnote ¹⁹ Before Constantine's acceptance of Christianity, moreover, Rome's religion was non-exclusive and did not seek to impose Roman gods and rituals on Rome's subjects, who were, with the notable exception of the Jews, also non-exclusive in their religions. Rather, a process of fusion was allowed to develop in which aspects of Roman religion were incorporated into the religious life of the provinces and vice versa. The late Victorians, by contrast, possessed highly developed pseudo-scientific theories of racial inequality, and while there was no overwhelming desire to convert the population of India, the British were not about to accept the religious beliefs of Hindus, Sikhs, and Muslims, as the Romans had those of their subjects. All three writers are forced to conclude that Roman-style cultural fusion could never take place, no matter how desirable it might seem as a way of retaining loyalty in the British empire.Footnote ²⁰

In addition, as Lord Cromer writes, there was a fundamental inconsistency in the motivations of the British which had never troubled the Romans:

LucasFootnote ²² and Bryce also recognize this paradox within the ideology of the British empire, the latter calling it ‘patent, but inevitable’.Footnote ²³ The Roman empire, in comparison, they notice, was never troubled with the ideals of national self-determination and self-government. Neither the central nor provincial governments possessed any representative institutions, and the goal was merely to rule and rule well. Half of the British empire, by contrast, consisted of the overtly titled self-governing dominions,Footnote ²⁴ which did indeed possess representative institutions similar to those of Britain itself. This compared uneasily with the reality of authoritarian rule in India. Even there, the ideal of self-government was always in the background. Indeed, Cromer argued that ‘the great imperial problem of the future [would be] to what extent some 350 millions of British subjects…aliens to us in race, religion, language…and customs, [were] to govern themselves, or…to be governed by us’. ‘Rome’, he admitted, ‘never had to face such an issue as this.’Footnote ²⁵

Yet the recognition of these important differences does not invalidate the comparison. What is striking about all three essays is the distinctly dispassionate tone in which they were written. In the words of a contemporary reviewer, they were attempting to conduct a ‘scientific inquiry into the problems of Imperialism’.Footnote ²⁶ What they were seeking in the comparison with Rome was not a model to follow, but a context, which would allow them to see the problems of their own empire more clearly, and perhaps even, on occasion, suggest possible courses of action, either to take, or to avoid. As Bryce writes in his opening pages, although ‘the conditions [in the two dominions] are in many respects different, yet there is in the parallel enough to make it instructive for the present, and possibly significant for the future’.Footnote ²⁷

A recognition of difference, even fundamental difference, does not invalidate the fact that in the task of contextualizing the problems faced by the British in their government of India, all three writers still considered the Roman empire the most appropriate comparison to make. The many other examples of empire provided by ancient and modern historyFootnote ²⁸ were rejected as bearing little significant resemblance to British India. ‘When we wish to examine the methods and the results of British rule in India by the light of any other dominion exercised under conditions even remotely similar’, writes Bryce, ‘it is to the Roman Empire of the centuries between Augustus and Honorius that we must go.’Footnote ²⁹ ‘Here surely’, appeals Cromer, ‘if it be true that history is philosophy teaching by example, some useful lessons are to be learnt.’Footnote ³⁰

Furthermore, this desire to contextualize and thereby to understand better modern imperial problems seem also to have motivated the efforts of several leading historians at this time who dwelt at length on the Roman empire and the intricacies of its provincial administration. As Bryce points out in the preface to his Studies in history and jurisprudence, the study of the Roman empire in his day no longer aimed simply to be ‘stimulative’. In the face of the modern problems of Indian government, it sought also to be ‘practically helpful’.Footnote ³¹ The Oxford ancient historian, Edward Fiddes, for one, echoed Bryce's sentiments on the potential value of Roman history for the better understanding of the problems of modern empire: ‘The Roman Empire’, he declared,

Moreover, there was an important belief in Britain in the late nineteenth and early twentieth centuries that an education in Roman imperial problems provided an excellent training for the rulers of British India. Augustin Filon, Arnold's French friend, believed that Roman history was ‘the best school’ for such men.Footnote ³³ Indeed, the study of Roman history at Oxford, had assumed, since its emergence in the 1880s under the aegis of Henry Francis Pelham, Camden Professor of Ancient History,Footnote ³⁴ an increasingly pragmatic and utilitarian approach. The teaching of undergraduates was considered much more important than research.Footnote ³⁵

The handbook, the epigraphical collection, and the historical commentary came to replace the great works of history, both narrative and theoretical, as the hallmark of Oxford scholarship.Footnote ³⁶ In a memoir of her brother, W. T. Arnold, Mary Ward remarks that he had been asked to compile a ‘Student's’ Roman empire, covering the period from the battle of Actium to the reign of Marcus Aurelius, one of a series of such ‘Student's’ manuals or handbooks which were in production at the time.Footnote ³⁷ Given that graduates of the Oxford Greats course provided nearly half of the successful candidates for the ICS Open Competition between 1892 and 1914, this emphasis on undergraduate teaching rather than academic research is understandable.

The prospect of practising real provincial government imbued the study of Roman provincial administration with a modern relevance. Indeed, it is the view of Oswyn Murray that ‘the best available education for a modern proconsul at the time was…close analysis of Pliny's Correspondence with Trajan’Footnote ³⁸ published for the purpose by E. G. Hardy in 1889. However, while the contemporary scholarly and reflective literature may reveal much about the popularity of the comparison with Rome, and, moreover, indicate the reasons behind its invocation, if we want to discover how important the analogy was as part of the education of the rulers of British India between 1890 and 1914, the most important source material to consult is the ICS Open Competition examination papers themselves. After all, their content dictated, to a very large extent, the education of candidates for the powerful Indian bureaucracy. A close analysis of examination questions reveals much about the ways in which the Civil Service commissioners expected candidates to think about British India and its government. First, it is important to note that just as Roman history reached a new level of influence in British academic circles during this period, so it did within the context of the ICS Open Competition. From 1892, Roman History appeared for the first time as a separate subject for examination and, moreover, there was a pronounced interest in the practical workings of the Roman empire. In 1894, for instance, candidates were asked to ‘Illustrate and explain the action of the Romans in dealing with problems of local government over the East and Western portions of the Empire respectively.’Footnote ³⁹

Significantly, for ICS candidates, the challenges surrounding the establishment of an effective imperial civil service within the Roman empire were a recurrent theme. In the 1911 Roman History paper, candidates were asked, ‘What were the chief hindrances in the formation of a regular civil service under the Republic, and what were the chief steps taken by Emperors to remedy the defect?’Footnote ⁴⁰

Between 1892 and 1914, not a year went by without at least one question being set on provincial administration in the Roman History section. This focus on the peculiarities of local government and its challenges was particularly appropriate for the examination of prospective ICS men, for a major part of their time as district officers in India would be spent dealing with local problems in their own particular area of the subcontinent. In the 1893 examination, candidates were asked explicitly, ‘What lessons with regard to the principles which should be observed in the government of British India may be learned from the history of the Roman dominions under the Republic and the Empire?’Footnote ⁴¹ And again, in the Roman History section of the 1912 examination, the question was posed, ‘How far is there any parallel in the problems of frontier defence between the Roman and the British Empires?’Footnote ⁴² When ICS officials were on leave, they would often discuss the lessons of Roman provincial policy for Indian frontier policy with W. T. Arnold.Footnote ⁴³

The ICS was responsible for the smooth running of British government in India and had been transferred to the direct control of the British crown (from the East India Company) following the Indian Rebellion of 1857, which had graphically demonstrated the challenges of maintaining control of the subcontinent. Given this history, it is not surprising that the commissioners should have been keen to get ICS candidates to think about how an imperial power should seek to maintain loyalty within its dominions. Indeed, the examination questions in the Roman History section reveal a persistent interest in the rebellions which took place in the Roman empire, especially in the grievances which lay behind them. In 1911, for example, candidates were asked to estimate the degree to which the Pictish chief Calgacus was justified in making his famous denouncement of Roman imperialism: ‘ubi solitudinem faciunt pacem appellant’ (where they create a barren wilderness, they call it peace).Footnote ⁴⁴ It was thus quite acceptable, desirable even, for a candidate to criticize Roman policy if the evidence called for it. More than anything else, the commissioners wanted future members of the ICS to be flexible enough in their thinking to adapt to whatever situations they might face in India, rather than being constrained by a particular way of thinking about empire and its problems.

II

While it is clear that the study of Roman imperial history was considered a useful intellectual tool in the education of the rulers of British India, precisely how important a tool was it thought to be? At a time when Political Science was held to be the ‘playground of analogies’,Footnote ⁴⁵ were other comparisons invoked which were similarly designed to stimulate thinking on imperial matters and prepare successful ICS candidates for the realities of governing an empire? And if so, what was the relative importance of the Roman comparison?

The year 1892, in which Roman History appeared for the first time in the ICS Open Competition as a separate paper, witnessed a revival of interest in ancient history in general, rather than in Roman history specifically. Greek History, too, made its debut as a separate paper. Moreover, the Civil Service commissioners showed as strong an interest in the organization and government of the empires of the ancient Greek world as they did in those of the Roman empire. In particular, there were questions set almost every year between 1892 and 1914 on the Athenian empire of the fifth century BC.Footnote ⁴⁶ In 1904, for example, candidates taking the Greek History paper were asked, ‘What is known as to the organisation of the Athenian Empire and from what sources is our information derived?’Footnote ⁴⁷ The same issues which were focused on by the commissioners in relation to Rome: the nature of internal ties, the grievances of subjects, and the action taken to tackle these grievances, were stressed in those questions dealing with the Athenian empire. In 1896, candidates were invited to analyse ‘the Athenian Empire, with respect to its internal organisation, and the nature of the ties which united the different members’.Footnote ⁴⁸ However, we must be careful when interpreting this evidence. At first sight, it may appear that Rome was but one of at least two examples of ancient empires which the commissioners felt it useful for candidates to consider in their preparation for careers as Indian civil servants. Yet, here, we should remember that Bryce, Lucas, and Cromer exercised caution when choosing which empire was most suitable for comparison with British India. Indeed, they rejected all except Rome as being too different to provide a worthwhile comparison with conditions in the Raj. The commissioners, though, were not concerned solely to encourage candidates to think about India apart from the rest of the British empire. Without doubt, India, with its huge population, cultural and religious diversity, and the direct manner of its government constituted a very distinct part of the empire, fundamentally different from its other parts, in particular the self-governing colonies. Moreover, the ICS was a unique body within the empire possessing unparalleled power and influence. Yet, for all this, the Raj was not outside the empire, but an integral part of it. Thus, while Bryce, Lucas, and Cromer were very much aware of India's uniqueness within the empire, they never sought to disconnect it from the larger whole. Lucas, for one, did not view the inconsistencies and fundamental differences in conditions between different parts of the empire as a problem. He called it ‘a unity of contradictions, absolutely impossible on paper, but working very comfortably in fact’.Footnote ⁴⁹

The commissioners shared this view. They sought not merely to educate their civil servants in the practical problems of administration, but also to encourage them to think about the wider issues of empire. Questions were regularly asked in the Political Science and Political Economy papers which encouraged candidates to consider the empire as a whole. In 1898, for example, the question was asked ‘What economic advantages does the United Kingdom derive from its colonies and dependencies?’Footnote ⁵⁰ Even though successful candidates would likely spend all their working lives in India, the commissioners expected them to take an interest in, and be able to analyse in some detail, the internal relations of all parts of the British empire. Thus, questions were regularly set which dealt with Britain's relations with its self-governing colonies. In 1901, for example, candidates were asked ‘How far does the present connexion between England and her Colonies seem to you permanent and what dangers threaten that connexion?’ As part of the same question, they were even asked to ‘suggest safeguards’ designed to hold the empire together.Footnote ⁵¹

Moreover, for these questions dealing with the wider empire outside India, particularly those concerning the relationship between Britain and its self-governing dominions, the Roman empire, was not at all helpful. In the words of Cromer: ‘no close or instructive analogy can be established between Rome in its relations with the provincials and Great Britain in its relations with the self-governing colonies’.Footnote ⁵² None of the provinces of the Roman empire had possessed any measure of self-government. The ICS examiners recognized this too, asking candidates in the English Law paper of 1910 to examine the statement:

Here, the Roman model was explicitly rejected in relation to the self-governing colonies. More appropriate analogies were needed, and were found by the commissioners, above all, in the ancient Greek world, particularly in the example of the Athenian empire or Delian League of the fifth century BC.

Although far from being a perfect analogy, conditions in the Athenian empire at least offered significantly more points of similarity, when compared with Britain's relations with its self-governing colonies. Indeed, to many, the loose links between the Greek city-state and its overseas settlements, which possessed a large degree of self-government, offered an attractive and potentially instructive parallel to the relationship between Britain and what Lucas termed its ‘sphere of settlement’. ‘There was’, he wrote,

E. A. Freeman, Regius Professor of Modern History at Oxford from 1884 to 1892, for one, interpreted the destiny of that portion of the empire which consisted of Britain and the self-governing colonies very much in the light of the confederate Athenian empire, and hoped that it would increasingly become a similar ‘unity of scattered kinsfolk’.Footnote ⁵⁵

The Civil Service commissioners picked up on the analogy too. Considerable interest was shown in asking ICS candidates to analyse the nature of Greek colonization, which Lucas, among others, had declared most suitable for comparison with the British in the self-governing colonies.Footnote ⁵⁶ Hardly a year went by without candidates being asked to: ‘Estimate the importance of the tie between a colony and its metropolis as a factor in Greek politics.’Footnote ⁵⁷ Above all, the commissioners were interested in getting candidates to think about and, more specifically, to assess the merits and demerits of the Greek principle of imperial federation, most famously displayed in the relations between Athens and its various subject-allies.Footnote ⁵⁸ In 1904, for instance, candidates sitting the Greek History paper were asked ‘What is meant by Federation? How far were its principles realised in the various unions of the Greek states in the fifth and fourth centuries B.C.?’Footnote ⁵⁹ The reasons behind this interest in Greek federal unions become clear when we notice that a parallel interest was shown by the commissioners in the principle of federation in the Political Science section of the examination.

Thus, in 1907, candidates were asked to ‘Explain and illustrate, either from ancient or modern examples, three of the peculiar weaknesses of federalism.’Footnote ⁶⁰ This question is a good example of how knowledge of the ancient world, gained in classical studies at university, could be of use to candidates outside the classical sections of the ICS examination. In Britain, from the 1880s onwards, the idea of an Imperial Federation of the British empire became increasingly influential. There was an Imperial Federation League, which had Lord Bryce as the chairman of its Oxford branch. Whilst the idea of an Imperial Federation had first been viewed only as an option for the peoples of Britain and the self-governing colonies,Footnote ⁶¹ it was clear that India and the other directly governed dependencies must somehow be included if the empire was to remain a viable single entity. Indeed, the question of what exactly the position of India and the other directly governed parts of the empire should be in such a federation became one of the most hotly debated political questions of the day. It is this debate which is reflected in ICS examination questions like those quoted above. Indeed, in 1904, the connection was made explicit, when candidates were asked to ‘Examine the conditions necessary to the successful working of a Federal government, and estimate the special difficulties that would present themselves in the case of a proposed Federation of the British Empire.’Footnote ⁶²

Thus, it seems clear that leading British statesmen-administrators and the Civil Service commissioners were intrigued with the potential lessons which might be learnt from political comparisons, but were not interested in forcing similarities. As Lord Cromer put it in a 1910 speech to the Oxford branch of the Classical Association, ‘Whilst it would be unwise to neglect the teaching of history, it is essential to steer clear of what is perhaps even a greater danger – namely that of being ensnared by the pitfall of false analogies.’Footnote ⁶³ This attitude is equally evident in the questions of the ICS examination papers themselves. In 1913, the commissioners invited candidates to consider and illustrate the statement: ‘It has been said that similarity in customs, laws and procedure, even in particulars, is only a guide to inquiry – that it is not proof of dependence or of a common origin’ and then to ‘briefly indicate precautions to be observed in the study of comparative politics’.Footnote ⁶⁴

The focus of the commissioners’ attention, when it came to analogies for thinking about the governing of British India and the empire more widely, was clearly on ancient empires. We have seen this not only in the Ancient History papers but also, and perhaps more tellingly, in the Political Science, Political Economy, and English Law papers. It would be wrong, however, to say that the commissioners showed no interest in asking candidates about more modern empires. The main difference to note here, though, is that such questions were exclusively confined to General Modern History papers and did not cross over into broader Political Science, Political Economy, and English Law papers. There were, on occasion, in the General Modern History paper, questions asking for comparisons among all of Europe's colonial powers. In 1905, candidates were asked ‘What were the chief relations of the different Powers of Europe with India in the latter half of the seventeenth century?’Footnote ⁶⁵

France, in particular, comes up quite frequently in questions about the British in India. Some are straightforward comparisons asking about methods of government which suggest that the commissioners saw enough similarity between the French and British empires in South-East Asia to make the comparison instructive. In 1914, in the General Modern History paper, candidates were asked to ‘Compare the treatment of their Asiatic possessions by England and France during the portion of the eighteenth century which is included in your period.’Footnote ⁶⁶ Likewise, in 1898, candidates were asked to compare Wellington and Napoleon as empire builders in India: ‘Lord Wellesley revived and proclaimed the Imperial principle of political supremacy. All his views and measures pointed towards the reconstruction of another empire in India. Compare and contrast Lord Wellesley with his great contemporary Napoleon Bonaparte.’Footnote ⁶⁷

III

Having considered the importance of Roman history and the Roman comparison in the training of Indian civil servants to think constructively about the practical problems of imperial government, we now turn to the other equally, if not more important, side of their preparation for an imperial career: their moral education. It was not enough that Indian civil servants should be able to think practically about imperial problems; it was even more important that they be the right sort of men. Many commentators in the late nineteenth and early twentieth centuries expressed the view that the British held India primarily by the force of moral example. In the words of Brigadier-General John Jacob:

Sir Charles Lucas, for one, shared this view. He argued that the distinctive features of the British character, which he saw above all as honesty and the instinct of fair play, contributed much to Britain's task of ruling the Raj. He wrote that Britain's success in holding India had been ‘mainly due to the combination of the strong hand with honesty and justice’. Britain had, he argued, ‘given [the Indians] from without what they had never received from within, security for life and property, justice between man and man, immunity from extortion, law instead of caprice’.Footnote ⁶⁹ Bryce too believed that in India the British had triumphed by force of character, and that it was primarily through their display of duty and moral courage that they continued to hold it.Footnote ⁷⁰ As R. R. W. Lingen, a senior civil servant of the time, explained: ‘the first and cardinal quality’ desirable in an Indian civil servant was ‘a deep self-sacrificing sense of duty; not a conventional one, satisfied by doing and avoiding certain prescribed or proscribed things, but such as is only satisfied when it has done its best, whether the sacrifice demanded thereby be that of ease, personal feeling, or private opinion.’Footnote ⁷¹ A high premium was therefore placed on finding men with the right moral training and qualities of ‘character’. But how important was the example provided by the Roman proconsul considered to be in this task? As we have seen, Edward Fiddes, for one, credited the Roman empire with being ‘the first great imperial experiment which rose above the methods of brute force…[and] made a genuine effort to unite Liberty and Empire’.Footnote ⁷²

At the start of his essay, Bryce too maintains that the Romans set themselves a ‘high ideal of the duties of a ruler’, which included the ‘responsibility for securing the welfare and the contentment of the provinces’.Footnote ⁷³ However, this does not seem to have been a view that was shared by most contemporaries. Indeed, Bryce is criticized in The Times review of his essay for ‘not recognising sufficiently…the fact that a higher standard of duties has been promulgated and in the main observed by our rulers [in India] than was ever enunciated with much effect at Rome’.Footnote ⁷⁴ The main problem with the Roman comparison here is that although the Romans may well have aspired to high standards in the government of their empire, they were generally seen to have failed in meeting them. Although Cromer credits Rome with at least a few ‘traces of the existence of a humanitarian policy’, visible in actions such as the Emperor Claudius's suppression of the perceived cruelties of Druidical worship, he is forced to conclude that ‘Roman imperial policy…if judged by such modern standards as we are wont to apply, stands condemned.’Footnote ⁷⁵ Even Bryce, when he considers Roman morality in more depth, argues that the experience of governing a large empire had a deleterious effect upon Roman morals: ‘It was the influence upon [Rome] of the conditions which attached to her rule in the provinces that did most to destroy…the old simple and upright character of the Roman people.’Footnote ⁷⁶

On many occasions, Bryce, Lucas, and Cromer emphasize what they perceive as the higher moral standards and more enlightened intentions of the British. Bryce argues that much of the Roman empire had been conquered because the Romans ‘enjoyed fighting for its own sake’ and ‘were content with slight pretexts for it’. By contrast, he argues, ‘The English went to India as traders…with no intention of fighting anybody.’Footnote ⁷⁷ At one point, Lucas stresses even more clearly the difference in intentions. He argues that ‘the ultimate aim’ of the Romans in the provinces was to ‘secure and maintain their tenure’, and that any beneficial ‘development’ which took place was a mere ‘by-product’ of that aim. The British, he stresses, were interested in holding India as much for the ‘benefit of India and the Indians’ as for themselves.Footnote ⁷⁸ Indeed, in the later nineteenth century, moral comparisons with the Roman empire were invariably drawn only to demonstrate the extent to which Britain had surpassed its ancient predecessor. People saw the British empire as democratic, driven by the high ideals of Christianity, and, as such, to have considerably improved upon the achievements of Rome.

The evidence of ICS examination papers reveals a similar picture. Nowhere is Rome held up as an example of moral imperial rule worthy of emulation by the British. Rather, questions often focused on rebellion and discontent within the empire and included the charge that the so-called pax Romana was a mere by-word for desolation.Footnote ⁷⁹ Yet, an analysis of the Moral Philosophy and Political Science sections of the examination papers does reveal at least one persistent trend: a profound interest in the moral and ethical qualities highlighted in the philosophy of Plato. For example, in 1896, candidates were asked to ‘Discuss Plato's views on the nature of virtue and the qualities held to be virtuous.’Footnote ⁸⁰

In particular, commissioners wanted candidates to consider the relevance of the Platonic virtues to the practice of government in the modern world. Thus, in 1900, candidates were invited to ‘Consider how far Plato's views of (a) the individual's duty to the State, (b) the State's duty to the individual, are applicable to modern life.’Footnote ⁸¹ Undoubtedly, the Platonic virtues of temperance, justice, courage, and wisdom were well suited to the needs of imperial administration. The commissioners were specifically interested in getting candidates to consider in detail the moral education which Plato prescribed for his ideal rulers, outlined in his Republic. In 1901, for example, candidates were invited to ‘Discuss the Platonic education as the ideal education for producing (a) a statesman, and (b) philosopher, and (c) a man of fully developed moral character.’Footnote ⁸² Plato argued that the prime moral quality desirable was a sense of selflessness and a total dedication to the welfare of those they ruled. They must, he wrote, ‘[n]either consider [n]or enjoin what is in [their] own interests, but always what is for the interest of [their] subjects’.Footnote ⁸³ Here, surely, was an inspiring example to set before the prospective ICS men whose most important quality was expected to be a ‘deep self-sacrificing sense of duty’.Footnote ⁸⁴ Philip Mason makes the connection explicit in his semi-official history of the ICS, when he deliberately likens India's civil servants to Plato's ideal rulers.Footnote ⁸⁵

Above all, Plato was interpreted by the late Victorians as arguing that government should rest in the hands of men who have received a broad and liberal education, rather than in those of technical experts. We see this view expressed across British society at the time, not just in the ICS but in the home civil service as well as in the public schools and universities.Footnote ⁸⁶ In Plato's prescriptions for the education of his specially trained Guardians, set down in the Republic, to quote R. M. Ogilvie, ‘the man who has been taught to rule merges into the broadly educated man’.Footnote ⁸⁷ In his own words, Plato's ideal rulers would have been brought up ‘from their youth’ to ‘imitate only those characters which are suitable to their profession – the courageous, temperate, holy, free’.Footnote ⁸⁸ The Guardians were to be devoted students of art, music, gymnastics, and, above all, philosophy. For, in Plato's view, any ruler ignorant of the pedagogic insights of philosophy, who has never contemplated the beauty of the world, and the intricacies of human nature, nor wrestled with the eternal problems of existence, is fundamentally ‘uncivilised…like a wild beast, all violence and fierceness, and knows no other way of dealing…he lives in all ignorance and evil conditions, and has no sense of propriety and grace’.Footnote ⁸⁹ Given the sustained interest in the Platonic ideals of higher education on the part of the commissioners in this period, it is hardly surprising that they desired their own candidates to have experienced a broadly based education. Indeed, the 1893 ICS examination syllabus declares that ‘the object of the Competition should be to secure for the ICS officers who have received the best, the most liberal, the most finished education’ available.Footnote ⁹⁰

This was generally felt to be provided by a thoroughgoing classical education, in particular, the Oxford Greats course, which, as we have already noted, provided nearly half of the successful ICS candidates between 1892 and 1914. Referring to the potential for success in the ICS Open Competition of the various candidates, J. L. Strachan-Davidson, then master of Balliol, wrote in a 1907 letter to Sir Arthur Godley, the permanent under-secretary of state between 1883 and 1909: ‘The Mathematical or Natural Science man has nothing to compare with the general training of the Classical scholar.’Footnote ⁹¹ Similarly, when questioned by the 1913 Royal Commission on Public Services in India as to whether men intended for the ICS should study anthropology and economics at Oxford as part of a specialized Honours School of Indian Studies, he replied that ‘if a man had a good education he would have learnt how to learn these things later’.Footnote ⁹² By a ‘good education’, he clearly meant Greats.

The Honours School of Classical Moderations and Greats seemed, in many ways, to offer an education for civil servants very much along Platonic lines. Primacy of place was given to the study of philosophy, in which the works of Plato, and of his pupil Aristotle, reigned supreme. Thus, students would meet Plato's plans for the education of the Guardians first-hand and could not fail to appreciate the similarities between them and their own philosophic education. There was, at Oxford in particular, a strong sense that it was the duty of the University to provide suitably qualified servants for the state. Strachan-Davidson, for one, was proud of the fact that Oxford, and its Greats course, provided the ICS with its ‘best recruiting ground’.Footnote ⁹³ Likewise, several Greats tutorsFootnote ⁹⁴ who supported Strachan-Davidson in his efforts to keep the ICS as a graduate service, declared to Lord Crewe, the secretary of state for India from 1910 to 1914, in a letter dated 7 March 1911, ‘Those who sign this letter have always laboured that Oxford should give of its best to the service of the country.’Footnote ⁹⁵ It is the opinion of Philip Mason that in our period the men of the ICS ‘were chosen and trained on Plato's principles as Guardians who would rule in the light of their own vision of the Good and the Beautiful’.Footnote ⁹⁶ And the Oxford Greats course, with its broad approach and its particular emphasis on philosophy certainly went some way to bringing these principles to life.

However, as the British were aware, no model was perfect. In some ways, the ideal of the Platonic Guardians was as limited in value for the moral education of ICS men as the Roman analogy was for the purposes of learning lessons in the practical business of empire. ‘The first main task’ of a Guardian, writes Philip Mason, ‘was to preserve order, to keep chaos at bay’.Footnote ⁹⁷ Plato's Guardians were at bottom despots, selfless and dedicated maybe, but still despots. They were a class apart, separated from the rest of society, chosen and educated specifically for the task of ruling, a task which could be given to no one else. While the lifestyle and ethos of Indian civil servants, with their elite sobriquet of ‘the heaven-born’ and their tendency to live apart, not only from the Indian population, but also from other Britons in India,Footnote ⁹⁸ may, in some ways, have resembled those of the Guardians, few ICS men appear to have seen their rule as permanent. Indeed, the ideals of trusteeship and eventual self-government for India were still held by many Britons in the late nineteenth and early twentieth centuries. Lord Cromer, for one, called self-government ‘unquestionably the most salient and generally accepted of our principles’, and stated his belief that, albeit in a carefully planned way, it ‘must manifestly constitute the corner-stone of the new edifice’ of India which the British were in the process of helping to construct.Footnote ⁹⁹

The commissioners were not unaware of these limitations. As with their treatment of the Roman analogy, they encouraged candidates to think critically about the questions set. Thus, while a clear interest was shown in Plato's recommendations for the education of his Guardians, in particular their value as a form of moral training for contemporary statesmen, nowhere were they propounded as constituting the best system. Rather, challenging questions were asked which demanded close and critical analysis of the thinker and his ideas. In 1904, the question was posed ‘Why were philosophers to rule in Plato's Republic? Has the rule of philosophers been successful so far as tried in practice? Refer to examples. What is the true function of philosophers in the Modern State, and what is the real and necessary relation of philosophers to the government?’Footnote ¹⁰⁰ Nor were Plato's ideas always considered in isolation. Frequently, candidates were asked to compare them with other moral and ethical systems which included both ancient and modern examples. In 1914, candidates were invited to ‘Discuss the relation of Plato's Moral Utopia to that of Herbert Spencer.’Footnote ¹⁰¹ British commentators in the late nineteenth and early twentieth centuries may have been deeply influenced in their choice of analogies by the Classics, but they were by no means limited to them alone. They were quite capable of looking outside their editions of Caesar, Plato, and Thucydides for analogies. Indeed, in 1895, ICS candidates were deliberately asked to confront the contention that on many occasions ‘the conclusions of ancient thinkers seem inapplicable to modern societies’.Footnote ¹⁰²

IV

Clearly, the Civil Service commissioners in our period showed a significant interest in ancient analogies, in using them to encourage candidates to think constructively about both the practical and moral issues of empire. What remains to be seen is to what extent there existed a tangible bias in the formal structure of the ICS Open Competition, in the papers set, and in the weighting of marks towards Classical subjects such as ancient history and philosophy from which these analogies were gleaned. The formal structure of the Open Competition underwent substantial reworking in 1892, and to a lesser extent, at two further points in our period: 1906 and 1911. Does an investigation of these changes reveal a deliberate attempt to structure the examination in favour of Classics,Footnote ¹⁰³ to the detriment of other subject areas?

The year 1892 witnessed major changes to the ICS's recruitment policy. The maximum age limit for sitting the Open Competition was raised significantly from nineteen to twenty-three, thus opening the ICS to university graduates. Moreover, a major change took place regarding the choice of papers included in the examination. The number of papers dealing with the ancient world, with its thought and civilization, increased dramatically. In addition to papers on the Greek and Latin languages and Logic, new papers were set in Moral Philosophy, Roman Law, and most importantly, Greek and Roman History, which also appeared for the first time in the ICS syllabus as separate papers. The subject areas of Mathematics, English, and Natural Science were enlarged too, but on nowhere near the scale of Classics: Mathematics was separated into two sections: standard and advanced level, English, into Literature and Composition, and Natural Science was sub-divided into seven smaller sections.Footnote ¹⁰⁴ However, candidates were limited to taking only three of these. In addition, completely new subject areas were introduced by the commissioners. Papers in English History and General Modern History were set for the first time, as well as in Political Science and English Law.

As we might expect, these developments were accompanied by a substantial readjustment in the allocation of marks (see Table 1). The introduction of four completely new subject areas in 1892 (Political Science, English Law, English History, and General Modern History), and the allocation to them collectively of some 15.6 per cent of the total marks available necessitated cutbacks elsewhere. The greatest casualty was Modern Languages, previously the most valuable subject area after Classics, worth 19.7 per cent of the total available marks. After 1892, it was worth only 7.9 per cent, a loss of nearly 12 per cent. English, too, suffered a loss of importance, going down from 12.7 per cent before 1892 to just 7.9 per cent afterwards. Mathematics did not lose any value but stayed virtually unchanged. It accounted for 14.2 per cent of the total available marks after 1892, compared with 14.1 per cent before.

Table 1. Weighting of different subject areas as a percentage of overall marks 1880–91 and 1892–1905

Source: A database compiled by the author deriving from the ICS Open Competition examination papers.

The treatment of Classics, however, was somewhat different. Not only, as already noted, were the number of its component papers increased significantly. Its overall proportion of marks was raised also. In total, after the changes of 1892, Classics was worth 3,600 marks, some 28.3 per cent of the total 12,700 available. This may only be a small increase of 4.4 per cent, but it is significant when it is compared with the corresponding changes in the marks allocated to the other main subject areas. Indeed, out of what may be called the four major subject areas in the ICS Open Competition: Classics, Mathematics, Modern Languages, and English (each of which accounted for more than 15 per cent of the individual overall scores of the top twenty candidates each year between 1880 and 1891), only Classics saw any real increase. The others either remained the same or were reduced in value. Moreover, a consideration of the examination results themselves reveals the gain made by Classics after 1892 to be even greater than the changes in the allocation of marks might suggest. Although, between 1892 and 1905, marks achieved in Classics made up, on average, only 3.2 per cent more of the individual overall scores of the top twenty candidates each year than in the decade prior to the changes (see Table 2), far more important is the fact that the Open Competition results tables also reveal a concomitant fall-off in the importance of marks earned in the other major subject areas.

Table 2. Importance of different subject areas in the individual overall scores of the top twenty ICS candidates each year 1880–91 and 1892–1905

Source: See Table 1.

Between 1880 and 1891, marks earned in Mathematics, for example, had constituted, on average, 22.5 per cent of the individual overall scores of the top twenty candidates each year. Between 1892 and 1905, however, it made up just 8.1 per cent, a loss of some 14.4 per cent. The most dramatic loss of importance in the scores of individual candidates, however, was suffered by Modern Languages. Between 1880 and 1891, it had, on average, accounted for 19.3 per cent of the individual overall scores of the top twenty candidates each year. Between 1892 and 1905, it made up just 4.7 per cent, a loss of some 14.6 per cent. The importance of English remained virtually unchanged. After 1892, then, the majority of the top twenty candidates each year specialized primarily in Classics rather than in the other major subject areas of Mathematics and Modern Languages.

After 1892, there were no more innovations regarding the actual choice of papers set for examination, but what further changes there were to the formal structure of the Open Competition only served to increase further the dominance of Classics. In 1906, at the urging of Strachan-Davidson, in particular, a maximum limit was introduced regarding the number of marks which candidates could score.Footnote ¹⁰⁵ Henceforward, candidates could no longer offer as many papers as they wanted, but only as many, which, when combined carried no more than 6,000 marks.

This measure was accompanied by a further readjustment to the allocation of marks. Arguing that since the changes of 1892 and the transformation of the ICS into a graduate service, Oxford Greats men had constituted the Service's ‘best recruiting ground’, Strachan-Davidson and his supporters urged the Civil Service commissioners to weight the marks even more heavily in favour of Classics. Despite the distinct classical bias already apparent in the changes of 1892, Strachan-Davidson argued that Greats men still had to go outside their degree subjects in order to prepare for the ICS Open Competition, and that this additional instruction, usually to be had with the notorious ‘crammer’,Footnote ¹⁰⁶ posed a serious threat to a candidate's potential for success in his university Honours degree.Footnote ¹⁰⁷

Consequently, as Table 3 shows, the number of marks available for Classics was raised to 4,500, which made up no less than 75 per cent of the 6,000 marks needed. Bearing in mind that on average the top mark scored by candidates between 1906 and 1910 was just 3,672 marks, and also the fact that no paper was compulsory, Greats men would no longer have had any need to go outside their university subjects of Greek and Latin, Ancient History, and Philosophy to be successful in the ICS Open Competition.

Table 3. Weighting of different subject areas as a percentage of overall marks 1906–10 and 1911–14

Source: See Table 1.

Furthermore, no other subject area had its number of marks raised so much, that, taken alone, it could enable a candidate to achieve the very highest scores. After 1906, the numbers of marks available for Mathematics and Natural Science were both raised to 2,400. Thus, a Cambridge Mathematics or Natural Science man would have been able to account for just 40 per cent of the 6,000 marks required with his university subject alone. Nor would he have been able to achieve the average top score of 3,672 marks without taking papers in additional subject areas. Worse still, a graduate of Modern Languages would have been able to account for only 30 per cent of the 6,000 marks with his university subject alone and would not even have been able to achieve 50 per cent of the average highest mark with it.

Moreover, an analysis of the examination results for the years following the changes of 1906 reveals that the importance of marks gained from Classics in the individual overall scores of the top twenty candidates each year continued to rise (see Table 4). Once more, it may only have been a small increase: between 1906 and 1910, on average, marks gained from Classics made up some 42.6 per cent of the individual overall scores of the top twenty candidates each year, compared with 39.4 per cent between 1892 and 1905. But it should be remembered that in both the officially orchestrated weighting of the marks, and in the actual examination results themselves, Classics was now, and had in effect been, since the changes of 1892, the dominant subject area in the ICS Open Competition. Before 1892, as we have seen, a total of four subject areas were each able to command on average over 15 per cent of the individual overall scores of the top twenty candidates each year.Footnote ¹⁰⁸ After 1892, only Classics and English continued to do so. Mathematics and Modern Languages were out of the running. English was the one subject area taken by all candidates from 1880 until 1914, and its average importance in the results of the top twenty candidates each year fluctuated only slightly during the entire period. And yet, after the further changes of 1906, an analysis of the results tables reveals that even English lost importance so that it accounted for less than 15 per cent of the individual overall scores of the top twenty candidates each year. After 1906, Classics stood alone and unchallenged.

Table 4. Importance of different subject areas in the individual overall scores of the top twenty ICS candidates each year 1906–10 and 1911–14

Source: See Table 1.

The changes of 1911, which were to be the final readjustments to the formal structure of the ICS Open Competition in our period, only made things easier still for Classics graduates sitting the examination. This time, it was only the number of marks allotted to Classics which was altered. And once more it was increased (see Table 3). Classics was now worth some 4,900 marks or 81.7 per cent of the required total of 6,000 marks, and there was a further rise in the importance of Classics as a proportion of the overall scores of the top twenty candidates, to 45.8 per cent (see Table 4).

Clearly then, from the time of the changes introduced in 1892, through the further alterations of 1906 and 1911, Classics was consistently favoured by the Civil Service commissioners at the expense of other subject areas. Indeed, Classics emerged as the only dominant subject area, both in the weighting of the marks and in the actual results of candidates.

V

There is no doubt that the tendency to compare the British and Roman empires in the period between 1890 and 1914 was a significant one. Nor was it restricted to broad comparisons designed merely to underscore Britain's identity as an imperial power in relation to its European rivals and the peoples over which it ruled. The comparison was made, not only by leading statesmen-administrators such as Bryce, Lucas, and Cromer, and historians like W. T. Arnold; with the introduction of the Roman History paper into the ICS Open Competition in 1892, it penetrated to the very heart of official thinking about the administration of British India. Candidates were expected to know in detail about the Roman empire and its practical problems. On occasion, candidates were explicitly asked to compare the two empires. It was believed that there was enough in the comparison with Rome to make it potentially valuable as a mine of information about imperial government and its enduring problems. Greek imperial analogies and ideas about empire were considered equally important, particularly for thinking about relations between Britain and the self-governing colonies. More interest was expressed in the moral training of Plato's Guardians than of Roman proconsuls, whose general standard of morality was held by many in Britain, including Bryce, Lucas, and Cromer, to have been of a considerably lower standard than that of the British imperialists of their own day. Clearly, what the commissioners were looking for in ICS candidates was a well-rounded classical training. They wanted men with a wide and varied knowledge of other civilizations which provided potentially valuable comparisons and analogies to the problems of contemporary empire.

While there is no doubt that this favouring of Classics worked to exclude those who had not enjoyed the benefits of an elite classical education, including, of course, many of the Indian candidates wanting to join the ICS, it was the particular value of the Classics as a field of reference, as a ‘playground of analogies’, which the commissioners sought to harness in the innovative changes of 1892. After the introduction of Greek and Roman History as separate papers in this year, neither was favoured over the other. The same number of marks was allotted to each. Indeed, if anything was particularly favoured in 1892, when the entry age was raised and the ICS was made a graduate service, it was the Oxford Greats course. The huge expansion of the classical part of the ICS syllabus in this year appears to have been deliberately designed to include the staple subjects of the Oxford Classical course: ancient history and moral philosophy.

By contrast, the changes would not have been greatly welcomed by students of Cambridge's Classical Tripos. In comparison with Oxford, Cambridge placed little emphasis on ancient history and philosophy. Its three-year degree was based far more on language and literature than Oxford's four-year course which packed linguistic and literary study into the first year, leaving the remaining three years for the detailed study of ancient history and thought. Ancient history and philosophy were the great storehouses of classical thought and knowledge containing centuries of practical imperial experience and reflections about the purpose and morality of empire which were precisely what the commissioners were interested in getting their candidates to think about.

We also considered the idea that, with its emphasis on the study of philosophy, Greats may have been seen as offering a modern version of Plato's ideal moral education for his Guardians. There is no doubt that Greats was valued for the type of man it was supposed to produce – a shrewd and independent thinker. In an Address to the Congress of the Universities of the Empire in 1912, Sir Stanley Leathes, who was First Civil Service commissioner from 1910 to 1927, and also a frequent correspondent of Strachan-Davidson, made the point clear. He argued that the combination of classical languages and history found in the Greats course was the ideal education for administrators in the empire as it taught young men how to think for themselves, and, moreover, that it taught them about human nature, a deep knowledge of which he considered the essential qualification for all forms of public service.Footnote ¹⁰⁹

Yet, it must be remembered that this special interest in Greats on the part of the Civil Service commissioners is a phenomenon with identifiable chronological boundaries. The Ancient History and Moral Philosophy papers, and with them the many questions we have considered dealing with aspects of ancient imperial experience, both practical and moral, were introduced at a specific point in time: 1892. As tempting as it might be to assume that classical imperial analogies were always influential in the formulation of British policy in India, this was not the case. It is true that from the earliest days of British expansion in the subcontinent comparisons were made between British activities and those of ancient imperialists, particularly the Romans.Footnote ¹¹⁰ However, until the late nineteenth century these were somewhat vague and unspecific. Ancient history did not come into its own as an academic discipline until the 1880s. Moreover, the comparisons that were made were not the result of an overtly utilitarian desire to improve the government of India. They were in many cases the largely disinterested reflections of men educated from early childhood in classical literature. After 1892, there was a different pragmatic tone involved in the drawing of classical comparisons, which is perhaps most clearly displayed in the works of Bryce, Lucas, and Cromer. What comes across in their essays is a desire to find new and better ways to govern India. As Cromer put it, ‘the foundations on which the British imperial policy of the future is to be based…are still in the process of being laid’.Footnote ¹¹¹