Chapter 9

Richardson’s Barometer: Colonial Representation in Grammatical Texts

Richard Steadman-Jones

The Manners of Mankind must ever form an interesting enquiry. In every age and climate they display a wonderful diversity of character; and exhibit a picture so variously coloured that we are convinced by experience alone, that the great original of the whole is Man. In all investigations of this important subject, Language claims a superior degree of attention: in so many points it will be found a most unerring guide; and, when viewed on philosophic ground, may be considered as the great barometer of the barbarity or civilization of a people.

John Richardson, 17771

It is now well known that the emergence of the British colonial state in late eighteenth-century Bengal prompted the servants of the East India Company to take an increased interest in the languages of South Asia. This new enthusiasm for linguistic study encompassed both the ‘classical languages’ of India, that is Persian and Sanskrit, and also some of its ‘vernaculars’, notably Urdu and Bengali. From the 1770s onwards more and more examples of that supremely didactic genre, the grammar, were published, each stressing the importance of a knowledge of languages to the effective administration of the Indian Empire.2 By the early decades of the nineteenth century so much material had appeared, at least on Urdu, that there was vitriolic competition among rival authors and publishers for a share of the growing market in grammatical texts and new recruits to the Company’s service would almost inevitably encounter such material in their early years in India. This essay will focus on one particular grammarian, a Scot named John Gilchrist, who arrived in India in the early 1780s and went on to publish a serialized dictionary of Urdu, complete by 1790, and a grammar of the same language in 1796.3

The last three decades have witnessed a wide-ranging, multi-disciplinary debate about the role of cultural activity, and particularly the ‘making of knowledge’, in the establishment and maintenance of colonial power. Given this interest in the political dimensions of intellectual activity, it is not surprising that a certain amount of attention has been paid to colonial grammars in general and the work of John Gilchrist in particular, notably by the anthropologist of South Asia, Bernard Cohn. In an important article on issues of language in British India, Cohn identifies complex political currents running through the material.4 On the one hand he sees the grammars as elements in a process through which western scholars transformed fields of knowledge which had previously been the province of Indian authorities, an argument with a distinctly Saidian flavour, although Cohn describes his analysis as ‘influenced by the work of Michel Foucault’ rather than aligning it explicitly with Said’s concept of ‘orientalism’.5 On the other hand he cites several writers, including Gilchrist, who emphasized the need to learn local languages in order to negotiate the minefield of Indian etiquette. Assertions of this kind betray a certain anxiety at the perils of conducting oneself within a hybrid colonial culture, an issue that, since the publication of Cohn’s article, has become an increasingly important area of investigation within the field of post-colonial theory.6

Cohn’s discussion of colonial grammars convincingly demonstrates that these rich and problematic texts have an important place in the emergent colonial culture of the late eighteenth century. His analyses, however, have tended to focus more on the programmatic statements which introduce and comment on grammatical publications than on the linguistic analyses presented in the bodies of the grammars themselves. The result is that, although they provide powerful accounts of the reasons the British had for learning Indian languages and the impact of colonial language policy on native speakers, they do not examine the ways in which the analysis of the forms and structures of languages could be understood as political. The aim of this essay is to enrich accounts such as Cohn’s by showing that embedded in the technical details of Gilchrist’s linguistic description are political statements in dialogue with those which appear in prefaces and appendices. This will be achieved through a close examination of the ways in which Gilchrist represents two particular features of the Urdu language. It may seem surprising to claim that technical grammatical discourse is susceptible to this kind of close reading but to eighteenth-century readers, for whom grammar was a basic component of education, the description as well as the deployment of languages was an inherently political activity.

This politicized view of language is beautifully articulated in the quotation at the beginning of the essay from the work of the eighteenth-century scholar of Arabic, John Richardson: ‘[I]n so many points [language] will be found a most unerring guide; and when viewed on philosophic ground, may be considered as the great barometer of the barbarity or civilization of a people’. This assertion rests upon the idea that western learning has successfully identified both the functions that language must fulfil in the life of a civilized people and also the forms that most effectively realize those functions. The nature of this idea will be explored more fully later in the essay. The important point here is that if a description of a language is felt to lay bare the character of a people, then it is easy to see how linguistic analyses could participate in an over-arching political rhetoric about the direction colonial society was to take. By presenting a language as more or less ‘civilized’, more or less-’barbarous’, a grammarian could advocate to his readership a particular way of viewing and interacting with its speakers.

This is the idea that will be explored in the rest of the essay and the argument will be articulated in three sections. The first will discuss the context in which Gilchrist’s texts were written and, with reference to Cohn, examine the anxieties that surrounded language learning in colonial India. The second will develop the reading of Richardson’s image of the barometer by exploring the ways in which linguistic analyses could produce different impressions of languages and their speakers. The final section will set out a detailed reading of two passages from Gilchrist’s work in order to illustrate the ways in which his linguistic representations interact with the agenda expressed elsewhere in his published writing. An important aspect of the discussion will be the ways in which commercial competition with other writers and publishers led Gilchrist to clarify the political meanings of his work. The need to win over his readership very much influenced the form of the texts that Gilchrist produced after his long years of study in the cities of the Gangetic plain and it will be important to understand the nature of the rivalries that he formed as he attempted to achieve this kind of communication.

The writer and his readers

John Gilchrist was born in Edinburgh in 1759, the son of a merchant named Walter Gilchrist.7 Despite this auspicious start, John’s prospects were ruined in the year of his birth by the disappearance of his father. Walter Gilchrist was rumoured to have moved to America or the West Indies but the family never succeeded in making contact with him again and ended up dispersed and in straitened circumstances. John was educated at George Heriot’s Hospital, a charitable institution for the orphaned sons of Edinburgh freemen, and was apprenticed to a surgeon at the age of fourteen. In the early 1780s he was serving as a surgeon’s mate in the Royal Navy and he seems also to have spent some time in the West Indies where, it is tempting to speculate, he may have been looking for this father. By 1782, however, he had decided on a change of course. He travelled to India seeking an appointment in the Company’s army and was rewarded with a position in General Goddard’s detachment, then stationed in Bombay.

Telling his story some sixteen years later, Gilchrist claimed that, as soon as he took up his appointment, it became evident that he needed to learn Urdu in order to communicate with the Indian soldiers or sepoys. A project that began as a practical response to his immediate circumstances, however, eventually took over Gilchrist’s life and transformed him from a surgeon’s mate into an orientalist scholar. There was little existing material for Europeans wishing to learn Urdu and Gilchrist resolved that he would fill that gap in the market. In 1785 he was given leave of absence to work full-time on a dictionary and grammar of Urdu and he travelled extensively in northern India in order to work with native speakers and to gather material. Between 1787 and 1790 the instalments of his dictionary were slowly and painfully published, and in 1796 he finally completed the undertaking he had begun over ten years earlier, by publishing his grammar of the Urdu language at the Chronicle Press in Calcutta.8

Throughout his published work Gilchrist emphasizes the practical benefits that would accrue to the Company if more of its servants spoke Urdu. Soldiers would be able to issue commands more effectively. Civil Servants would have a firmer grasp of the administrative proceedings they were supervising and would be in a better position to protect the interests of ordinary Indians from a supposedly corrupt elite. At the same time, however, Gilchrist always saw his linguistic work as a means of advancing his own personal fortunes. For the boy whose father deserted him in the year of his birth financial concerns were never far away and the prefaces to his published volumes frequently contain detailed discussions of the sales of his earlier works. Gilchrist’s linguistic writing needs to be interpreted in this commercial context. To ensure that his texts were widely read by the servants of the Company, he needed to establish that they were the best material available in that relatively small market and this involved fighting off one competitor in particular.

In 1772 Captain George Hadley, recently retired from the Company’s service in Bengal, had published an outline of Urdu grammar along with a vocabulary under the title, Grammatical Remarks.9 The work appeared in London. Hadley’s text did not describe the elevated styles of Urdu to which Gilchrist was exposed in his work among the ‘learned natives’ of the northern cities. It presented a simplified variety of a kind used in interactions with soldiers and servants. There is virtually no attempt to render Urdu sounds which are difficult for speakers of English, for example the retroflex flap and stops,10 and in a number of ways the grammar resembles that of a pidgin – for instance, in the use of independent lexical items to mark the plurality of nouns.11 Despite its failings, however, it is clear that there was a real demand for a work of this kind and Hadley’s grammar achieved a certain degree of popularity. A second edition appeared in 177 4 and a third in 1784, when Gilchrist was preparing for his leave of absence. Indeed when Gilchrist first started to learn Urdu he was ‘of course referred to Hadley for the rudiments of the language’ but he quickly formed a low opinion of the text and this motivated him to assemble materials of his own.

One of Gilchrist’s major concerns was to dissuade his readers from relying on the ‘jargon’ described in Hadley’s little book and to apply themselves to a variety sanctioned by the ‘learned natives’ with whom he himself had worked.12 Clearly this was a commercial strategy, at least in part. Hadley’s grammar had a healthy readership, which Gilchrist saw as a potential market for his own publication. Nevertheless, Gilchrist’s warnings against ‘jargon’ form part of a highly politicized account of life in the colony. Gilchrist was obsessed with the need to understand Indian codes of etiquette, not least because of the danger that they could be exploited to produce subde forms of anti-colonial resistance, a position well summarized by Cohn:

The European has to learn to insist on proper performance of the Indian’s social and verbal codes in dealing with superiors. One should not let an Indian subordinate get away with behaviour or speech acts which would be offensive not only to the European but to an Indian of superior quality.13

If the servants of the Company did not grasp local standards of behaviour they risked becoming objects of mockery or ridicule.

This argument extends to linguistic behaviour in various ways. In the grammar, for example, Gilchrist warns the reader to pay attention to the use of familiar and respectful pronouns in Urdu, in case he is unwittingly insulted by his interlocutor. He asserts that ‘details of this kind’ may be of ‘more consequence in our daily transactions with the Hindoostanees, than we have hitherto been aware of and warns that small errors will end in the ‘humiliation and extinction of the British name in this portion of their empire’.14 To use Hadley’s ‘jargon’ is similarly to invite ridicule and Gilchrist goes to some lengths to emphasize that Indians themselves have very little respect for it. The ‘jargon’, he says, ‘exists no where, but among the dregs of our servants, in their snip snap dialogues with us only’. Even the servants do not use it amongst themselves for ‘they would not degrade themselves by chattering the gibberish of savages, while conversing with, or addressing each other, in the capacity of human beings’.15

According to Gilchrist, however, Hadley’s grammar had convinced a whole generation that the pidgin variety used with servants exhausted the scope of the Urdu language and that there simply was no other variety to learn. Gilchrist later claimed that from the outset it had been his intention to dispel this myth and establish Urdu as an object of serious study. In retrospect, he was astonished at his own temerity:

[I was] a young man, who had hardly been three years in India, and yet dared to dissent from, and teach those wise heads, who had grown grey in the service, ‘that the Hindoostanee was a language, worthy their acquisition, in every respect, and moreover, that such a thing as a jargon being current over mighty civilized empires, was a monstrous conception, that could exist nowhere but in their own brains.’16

Of course Gilchrist cannot be characterized simply as the champion of ‘proper’ Urdu in opposition to the bastardized variety of the kitchen and the camp. As Cohn points out, the decisions that writers like Gilchrist took about which varieties to describe and in what manner to describe them to some degree transformed the languages themselves and hence also the social reality of which they were a part.17 Nevertheless, the idea that Urdu is a civilized language, to be taken seriously by the British community, is a central part of Gilchrist’s message and to communicate it effectively it was necessary for him to distance his work as much as possible from that of Hadley.

One way in which Gilchrist did this was to construct his work both in style and substance as the product of scholarship rather than simply of practical experience. The dictionary and grammar together form a series of three quarto volumes. The quality of the typography is high and there is a great quantity of Perso-Arabic typeface. The grammatical description is wide-ranging and detailed and each point is supported by copious quotations from the Urdu poets. There are also extensive footnotes, which refer frequendy to western authorities so that Gilchrist’s mastery of both eastern and western knowledge is asserted throughout the text. This carefully constructed edifice of scholarship certainly impressed the elite elements of the colonial establishment. When Gilchrist returned to Calcutta in 1798, Marquis Wellesley, the Governor-General, gave him the opportunity to run an ‘oriental seminary’ to give newly recruited servants of the Company a grounding in both Urdu and Persian. Then in 1801, when this experimental project was superseded by the opening of Fort William College in Calcutta, Gilchrist was made Professor of Hindustani and put in charge of one of the largest departments in the new institution, a position of considerable influence.18

Yet Gilchrist’s early works did not find such favour with the general readership. A fourth and expanded edition of Hadley’s work was published in 1796, the very year in which Gilchrist’s grammar appeared. This was certainly a canny move on the part of the publisher since the little book of ‘jargon’ continued to sell well and new editions appeared in 1801, 1804, and 1809, with some evidence of pirated editions also circulating in Calcutta.19 It is unsurprising that ordinary readers preferred a compact and straightforward octavo volume to Gilchrist’s monumental stack of books and in the preface to a new work of 1798, The Oriental Linguist, Gilchrist states that he has been experiencing ‘pecuniary embarrassments’ as a result of ‘[t]he sale of Hadley’s insigniftcant catch-penny production’.20 In order to compete he reorganized his material into a series of more ‘conciliating’ publications.21 The ferocious performance of scholarship is modifted; dialogues and familiar phrases are incorporated in imitation of some of Hadley’s later editions; the grammatical exposition is considerably simplifted; and Urdu forms are presented in Roman characters only.

Gilchrist’s later works abandon the strategy of establishing difference and compete with Hadley’s by imitating them in all but the language they describe. Throughout the full range of his work, however, Gilchrist continued to emphasize that the variety he was describing was not a demeaning ‘jargon’. A constant theme of his writing, whether he is playing the orientalist scholar or the popular language master, is that Urdu has all the marks of a ‘civilized’ language, the marks that in Richardson’s terms make the language a barometer of the condition of its speakers. By foregrounding features of the language which, for eighteenth-century readers, seemed ‘rational’ and ‘civilized’ and by the artful presentation of ones that might have seemed more problematic, he produces a representation of Urdu as a language worth studying properly and not a variety that even the servants disdain to use. As we address the question of how languages were felt to act as indicators of ‘civilization’ and ‘barbarity’ it will be important to remember the sense of disquiet which informs Gilchrist’s analysis. Read in this way, his artful deployment of the forms and structures of the Urdu language emerges as a warning that the colonizer scorns the culture of the colonized at some peril.

The politics of grammatical representation

Over the course of the nineteenth century the study of language was gradually reconceived as a science in which generalizations were built on the empirical investigation of data and linguists increasingly asserted their autonomy of the philosophical disciplines.22 During the eighteenth century, however, when Gilchrist and Richardson were writing, languages were usually characterized not as natural entities but as the products of human ingenuity or, to use the contemporary terminology, as ‘arts’ invented to solve the problem of externalizing thought and communicating it to others. It was the task of the practical grammarian to describe how a particular language solved this problem by reducing it to rules for the benefit of learners. The investigation of language as a human faculty, however, was conducted firmly under the auspices of other disciplines.

If language was an ‘art’ for the communication of thought, for example, then the nature of language should surely be investigated with reference to those other forms of inquiry which dealt with the nature of the mind and the world, logic and metaphysics. The field of ‘general grammar’ was concerned with just this kind of analysis. Linguistic facts, the existence of nouns and adjectives, for example, were explained with reference to categories provided by the various philosophical sub-disciplines, in this case the metaphysical distinction between substance and attribute. The ‘inventors’ of language were held to have created these two parts of speech because both kinds of entity exist in the external world and so human beings need to communicate ideas about both of them in ordinary discourse.23

It is important to emphasize that general grammar developed as a way of exploring western languages, particularly Latin, Greek, and French, and that it effectively constituted a philosophical justification of the forms of those particular languages. When travellers, traders, and colonists encountered non-European languages that did not conform to the same formal patterns, it was not always straightforward to assimilate them into this framework of thought. The fact that it was difficult to correlate the forms of ‘exotic’ languages with the categories of western philosophy in some cases led to the languages being criticized as ‘unphilosophical’ and ‘barbarous’. If philosophy provides universal truths about the nature of the world, the reasoning went, then languages which show no awareness of those universal truths must indeed be inferior to those which do.24

Richardson’s idea of language as ‘the barometer of the barbarity or civilization of a people’ rests upon this kind of reasoning. Languages are more or less ‘barbarous’, more or less ‘civilized’, to the extent that they show an awareness of the nature of the world as revealed in western learning. For a writer such as Gilchrist, striving to persuade his compatriots that they have to rethink their relationship with a language they perceive as a ‘barbarous jargon’, this perspective opens up certain problems. Even today, language learners are likely to scoff at features of languages that they find strange or difficult and, when this attitude was reinforced by a scholarly discourse that stigmatized such features as ‘barbarous’ or ‘unphilosophical’, the practical grammarian faced serious difficulties in attempting to change attitudes.

But the ‘knowledge’ produced within discourses like the ‘general grammar’ was not as monolithic as the preceding discussion suggests. It was open to grammarians like Gilchrist to criticize, to modify, and to interact with that body of material in a more creative way than is often thought possible within the unpromising genre of grammar. Moreover, it is the procedures that he used in negotiating this interaction which allowed him to undertake his ‘defence’ of Urdu. First, there might be features of the language, which could actually be presented as ‘philosophical’ in western terms and these could be foregrounded through rhetorical devices including metaphor and other kinds of imagery. Second, the European languages sometimes revealed themselves in surprising ways when subjected to close scrutiny and descriptions of non-European languages could be authorized by showing that they worked in the same way as some accepted analysis of English or Latin. This kind of procedure had the advantage of allowing the grammarian to cite the names of earlier scholars in his work and even suggest that he had some kind of intellectual community with them.

It is important to recognize that, as grammarians such as Gilchrist set out the forms and structures of languages, they inevitably opened them up for scrutiny and criticism. At the same time, however, through the artful deployment of material, it was possible for them to anticipate and avert this hostile attention. This will become evident as we examine two areas of the grammar where Gilchrist applies the kind of strategies outlined above: first, the treatment of irregular verbs, and second, the expression of the concept of possession. Given the importance that Gilchrist places on detail in the presentation of the colonial self, it is natural that the political meanings of his work should also be visible in these detailed linguistic analyses.

Signs of civilization: the verbal system

A feature of Urdu that Gilchrist consistendy singles out for praise is the uniformity of its verbal system. He stresses repeatedly that Urdu has only one conjugational pattern. Chapter five of the grammar, ‘Of Verbs’, begins: ‘We are now arrived at that part of the Hindoostanee, which of all others will prove the most agreeable and satisfactory; the whole of the verbs being reducible under one conjugation, whose changes of mood and tense … are as obvious and easy, as the personal inflections are simple and uniform.’25 In the preface to the dictionary, he repeats this point and says of the verbal system that no other part of the grammar ‘can afford so much to admire, and so little to censure’.26 He also emphasizes that Urdu has few irregular verbs. In the grammar he asserts that they are ‘so few in number, as scarcely to merit a recital; which is a circumstance that has not probably a parallel in any ancient or modern language, and will be very acceptable to the reader’.27

For the modern learner, too, these features are attractive since they render the acquisition of the language simpler. For Gilchrist, however, the uniformity and regularity of the verbal system have a deeper significance. Linguistic texts of the seventeenth and eighteenth centuries often express the idea that a truly ‘rational’ language would be entirely regular and uniform and that the more irregular a language is, the more defective it must be. In 1668, for example, Bishop John Wilkins, a leading member of the Royal Society, published a proposal for a ‘philosophical language’ that was to serve as a vehicle for scientific inquiry.28 Wilkins’ essay includes a critique of the grammar of Latin intended to show that this rival medium of learned discourse was in fact unsuited to the role that had been conferred on it. Wilkins criticizes the fact that Latin nouns are distributed across five declensions. ‘They are’, he says, ‘unnecessary and inconvenient’.29 Similarly the ‘four distinct ways of conjugating Verbs’ result in an absurd proliferation of forms.30 Wilkins’ own philosophical language will not suffer from these failings. Nor, according to Gilchrist, does Urdu. Thus, his repeated assertion that all the verbs are ‘reducible under one conjugation’ must be read in the light of the notion that reason should inform the construction of languages.

Yet the simplicity of the Urdu verbal system was not obvious to all eighteenth-century observers. In the first three editions of Hadley’s work the language is said to have no less than five conjugations.31 Thus, to establish the rationality of Urdu grammar, Gilchrist needed actively to argue against this position at a technical level. His argument proved persuasive and, in the fourth edition of Hadley’s work, only one conjugation is postulated. This change of heart, of course, elicited vociferous accusations of plagiarism from Gilchrist. The image he uses in vilifying Hadley is interesting. He compares the grammar of Urdu to a landscape and asserts that the verbal system is the most beautiful part of the ‘philological champaign’. Hadley, meanwhile, is compared to an owl ‘who has long perched on the watch tower of Hindoostanee grammar … with hootings ominous and fatal enough, to have far encroached on the long secluded day’.32 The point of the comparison is that the owl’s ‘nocturnal vision’ has prevented it till now from seeing the beauty of the landscape. Thus, Hadley’s failure to grasp the simplicity of the verbal system is understood as a failure to appreciate its excellence.

In his critique of Latin grammar, however, Wilkins also deals with the issue of irregularity. Latin morphology would be problematic enough, he says, if the rules were ‘general and constant’, but ‘the exceptions and Anomalisms … are so very numerous that there is much more pains required for the remembring [sic] of them, than of the Rules themselves: insomuch that many eminent Grammarians have written against Analogy, both in Greek and Latin’.33 He offers a ‘brief view’ of these ‘anomalisms’ stating that there are so many exceptions in the conjugations of verbs, particularly in the formations of preterites and supines,34 that it is difficult to list them. ‘Some are wholly without them, others have them without any Analogy; as Fleo Flevi, Sero Sevi, Fero Tuli.’ Furthermore, different verbs sometimes have the same preterites or supines while some are ‘of none of the four conjugations’: sum (I am), volo (I want), fio (I become), and eo (I go), for example.35

Urdu undeniably has some irregular verb forms, albeit few, and Gilchrist clearly sees them as imperfections. The perfect participles of karnâ (to do), marnâ (to die), denâ (to give), and lenâ (to take), for example, are irregular. We would expect their perfects to be karâ, marâ, dryâ, and leyâ and the fact that the forms kiyâ, muâ, diyâ, and liyâ are in use clearly needs explanation.36 In presenting each, Gilchrist provides a brief discussion of how he believes the anomaly to have arisen. In all four cases he claims that the problematic forms are not irregularly derived from contemporary infinitives but are regular formations from old or disused infinitives, namely ‘keena’, ‘moona’, ‘deena’ and ‘leena’.37 The same explanation appears in one of his later works, The Anti-Jargonist.

The ancient infinitives keena, moona, have probably left their regular preterites kee, a, moo, a as dying bequests to kuma, to do; murna, to die, and when proving too powerful mementos have almost buried in oblivion the regular tenses kura, mura, of these surviving legatees, who may thence be said to inherit rather superfluous than irregular preterites.38

Gilchrist goes on to say that this explanation will

in a great measure rescue the language of Hindoostan from the reproach of a single irregular verb, in this philological attempt of ours to recover its disfigured body, from the clumsy paws of those bruins, who have preposterously endeavoured to lick its mangled carcase as a jargon, into some intelligible shape or form.39

The implication is clearly that the existence of obsolete verbs from which the anomalous forms were derived would mitigate the anomalies, and the legal metaphors strongly reinforce the normative character of the discussion. Gilchrist is not merely showing how the irregular forms have come into existence but attempting to justify them by showing that they do not violate the principles on which languages should be constructed. It is also important to note that the project of ‘rescuing’ the language is set in opposition to the work of ‘bruins’ such as Hadley, whose attempts to analyse a ‘jargon’, by defmition, entail the acceptance of linguistic corruption. Here, then, Gilchrist pursues his communicative goals by presenting an argument from the history of the language and he reinforces his point by using metaphor. He implicitly likens himself to a lawyer and his opponents to the comical animals of fabular tradition, licking a corpse in an effort to make it live. In so doing he hammers home the message that Urdu is ‘rational’ and ‘civilized’ while at the same time representing himself, in contrast to Hadley, as a reputable scholar. By foregrounding a feature of the language that is identified within philosophical texts as a virtue, Gilchrist is able to comment upon the nature of Urdu, upon his own status as a scholar, and upon the deficiencies of his major rival, Hadley.

The possessive adjective

The second example of Gilchrist’s grammatical analysis relates to the postposition, kâ, which is used in Urdu to indicate possession. It is a feature of this particle that it must agree in gender and number not with the noun to which it is attached but with the following noun. Thus ‘Sunil’s brother’ would be translated sunîl kâ bhâî where kâ is masculine singular to agree with bhâî (brother). ‘Sunil’s sister’, however, would be translated sunîl kî bahn where kî is feminine singular to agree with bahn (sister). Within the parameters of eighteenth-century linguistic thought this might be viewed as strange behaviour for a postposition. Prepositions were categorized as belonging to the ‘undeclined’ parts of speech and having no ‘accidents’ to be expressed through inflection.40 Postpositions were understood to be similar in character to prepositions, so what is the justification for a postposition that declines?

Gilchrist’s approach is to claim that the particle kâ is actually an adjectival ending that converts the preceding noun into an adjective. This seems a more acceptable description. Eighteenth-century readers with even the most basic knowledge of French, Latin, or Greek would be perfectly used to the idea that adjectives have inflections. What might continue to cause problems, however, is the idea that possession is expressed by converting the noun marking the possessor into an adjective. Surely possession is a relation and should be expressed either by a case ending or a preposition? But this is another challenge to which Gilchrist can rise and he does so by drawing a comparison with a feature of English grammar, namely the apostrophe-s, which is also used to denote possession.

Gilchrist refers to a work by another member of the seventeenth-century Royal Society, John Wallis’ Grammatica linguce Anglicance, first published in 1652, and says, ‘This genetive [the postposition kâ], like ours [the apostrophe-s], so much resembles an acjjective, that I cannot help thinking the learned Wallis had some reason for calling this form of the noun one’. The reference is to Wallis’ chapter on adjectives where he claims that there are two types of adjectives that are formed directly from substantives: ‘The first of these may be called the possessive acjjective [Acjjectivum Possessivum]. It is formed from any substantive, singular or plural, by the addition of s (or es, if required by the pronunciation)’.41 This is a clever tactical move on Gilchrist’s part. If the learned Dr Wallis thinks that English expresses possession by converting nouns into adjectives then it is hardly appropriate to criticize Urdu for doing the same thing.

To a certain extent Gilchrist is interpreting the discussion of the acjjectivum possessivum for his own purposes. Wallis does not explicitly discuss his motives for describing the apostrophe-s form as an adjective but the anonymous author of the Bellum Grammaticale, a work published in 1712, which discusses three contemporary grammars of English, provides a suggestion. This author criticizes one of his targets, James Greenwood, for not following this analysis and says that, if it is adopted, ‘there is no need of endeavouring to thrust one Case into our Language, quite contrary to its Genius, which hates Cases, since the Way Dr. Wallis has judiciously chose, answers the End as well, and agrees better with the Nature of our Tongue’.42 On this reading, the argument is language-speciflc. In order to avoid saying that English has any cases at all, Wallis has made the one possible candidate into an adjectival termination.

Gilchrist, however, does not present Wallis’ analysis as a language-speciflc argument that happens to provide a model for his own interpretation of the Urdu data. Instead he treats Wallis’ analysis as if it were presenting a crosslinguistic theory about the expression of possession and he characterizes his own discovery concerning Urdu as providing further evidence for this universal theory. This impression is reinforced in one of his later works, The British Indian Monitor, where he states that ‘Ka, of, ’s has not only all the governing qualities of a postposition in the Hindoostanee, but is itself a declinable adjunct, that admirably proves the intimate connexion between genitive and adjective forms in most languages’.43 By this stage the theory has expanded to embrace ‘most languages’ and the claim is supported with the observation that the Latin form cujus (of whom, of which) is sometimes treated as the invariable genitive form of the relative pronoun and sometimes as an adjective which varies for gender, case and number. The implication is that in such illustrious company – first English, and now Latin – Urdu can hardly be criticized for its adjectival possessives.

In his discussion of the expression of possession Gilchrist takes a potentially problematic feature of Urdu and finds precedents for it in prestigious analyses of other languages. The very fact of citing Wallis’ important grammar of English is a powerful strategy of self-legitimation as a scholar. Instead of simply pointing to Wallis’ work in defence of his analysis, however, Gilchrist goes a stage further by putting the two analyses side by side and claiming to be developing a universal theory which originates with Wallis, a claim which becomes more strongly expressed as time passes: the similarity of adjectival and genitive forms in Urdu, English and Latin is evidence that, in some underlying sense, the genitive is an adjective. In this way Gilchrist presents himself as someone akin to Wallis, who thinks philosophically about language and is not, like Hadley, a mere purveyor of shoddy text books to undiscriminating readers. Again, in the midst of his technical grammatical description, he is developing a detailed representation of himself and his object of study.

Conclusion

The preservation of face was of primary importance for the Company servant who was presenting himself before the colonized population. In the institutional context of Fort William College Gilchrist was given authority to convey this message to students for whom attendance at the College was a requirement. To reach a more general readership, however, he had to compete in the literary market place and this involved producing texts that would win readers over from Hadley’s descriptions of ‘jargon’ and convince them that they needed to engage with the styles of Urdu presented in his own work. It is clear that the mass of scholarship presented in the grammar and dictionary did not serve this purpose. It impressed Wellesley enough to appoint Gilchrist to the staff of the new college, but it possibly alienated those other readers who continued to purchase Hadley’s little book throughout the first decade of the nineteenth century. As the style of the texts changed, however, and Gilchrist’s books began to resemble Hadley’s in their provision of dialogues and their more compressed treatment of the grammar, the kind of strategies described in the last section endured. In a sense it mattered little whether individual readers were interested in the theory of general grammar or had even heard of the learned Dr Wallis. What is important is the cumulative effect of repeatedly setting Urdu alongside European languages, finding it a match for them, and arguing for its learned and worthy qualities.

In the late eighteenth and early nineteenth centuries, colonial anxiety informed the construction of languages as objects of study at the most detailed level and the grammatical literature of empire is particularly interesting from this perspective because in it we see members of the British community attempting to mobilize opinion amongst themselves. In the late twentieth century grammar is no longer the foundation of a liberal education. However, in the eighteenth and nineteenth centuries grammatical studies were still central even to basic schooling. The result is that texts that would have seemed allusive, provocative, even witty or playful to their contemporary readers have hardened and been rendered opaque to the modern eye. We need to learn how to read these texts because they are not simply the repositories of out-of-date technical information. They are an important source of nuanced information about the micro-economy of cultural encounters in the period of emergent colonialism.