1. Foreword

This volume, entitled Early Western and Portuguese descriptors of the South Asian languages from the 16th century onwards, contains selected papers presented during the 32nd edition of the South Asian Languages Analysis Roundtable (SALA–32) held at the University of Lisbon, Portugal, on April 27th–29th 2016. The panel within which they were presented was entitled History of early linguistic descriptions of Indian languages, and it invited contributors to discuss the early linguistic works describing South Asian languages without limitation on the perspective or the framework to be selected by scholars in the analysis, discussion, and presentation of the history of early linguistic descriptions of South Asian languages.

At present, and respecting the same methodological criteria, the same scholars are involved in this editorial project with the aim of bringing together in a single volume those different perspectives which discuss and analyse the works produced by those who can be defined as linguists avant la lettre. Hence, the publication of the present volume, as well as the organisation of the panel within SALA–32, serves the specific purpose of contributing towards the advancement of scholarships on the knowledge about early descriptors and descriptions of South Asian languages by reflecting on the implication of the process of grammaticisation of non-European linguistic features and on different aspects of these linguistic works avant la lettre as well.

As has been largely described by Auroux (1992, 1994), the history of language description has always been characterized by grammaticisation, or rather that the activity of the description of languages and the elaboration of grammars.1 Auroux (1994) was pioneering in introducing the notion of grammaire étendue (extended grammar) through which he defined the dialogue between different grammatical traditions, highlighting both the borrowing as well as the adaptation and thus the extension of grammatical models used to describe languages different from those for which they had been originally elaborated.2

After the exploration era, the need for the acquisition of unknown languages led to the grammaticisation of these newly discovered languages, among which the South Asian ones which were mainly described through the lens of the Western grammatical framework (the Latin or the Greek grammar). The observations led to the discovery of linguistic structures which could not fit into the Western model of reference. For this reason, the original model needed to be extended, gradually contributing, in this way, to the development of the Science of Language. However, grammaticisation was much more than this. Through this, not only were languages described, but also the ‘new’ linguistic knowledge about the newly discovered languages was disseminated in Europe as well as in the territories where the linguistic works were produced, and its circulation lead to influence both the status of languages within the new communities, but also within the communities where these newly discovered languages were originally spoken.3

For this reason, and because the study of the history of the early descriptions can lead us to different paths of reflection and analysis, the main train of thought behind these papers has been left as general as possible. Indeed, it resides on the fact that all the papers here presented in a single volume deal with the Indian subcontinent; secondly, that they all discuss, analyse, or report on the activity of descriptors, collectors, and editors who contributed to the grammaticisation of South Asian languages as well as to the diffusion on a new linguistic knowledge, in the pre-modern period – the latter with consequences on the implementation and diffusion of ideologies in India toward specific South Asian languages (e.g. Annamalai 2011: 13–34).

Apart from this, the papers in the volume do not share other common topics, but rather their main contribution to the volume is represented by the different angles through which early descriptions of South Asian languages are framed and discussed, within both macro and micro perspectives.

First of all, three of the five papers deal with the Tamil language (Chevillard, James, Muru) and two with Indo-Aryan languages, Sanskrit and Konkani respectively (Ciotti, Fernandes). This imbalance between Tamil and other languages is mainly due to the fact that Tamil was one of the vernacular languages of which the early missionaries wanted to attain sufficient command of. For example, their interest in Sanskrit, representative of the sacred language of the Hindus, only appeared at a second stage for two main reasons, a strategic one consisting in the will of also converting Brahmins to Christianism, and a cultural one determined by an increasing interest in understanding the traditional culture and religion. Furthermore, many of the manuscripts that have been discovered so far and that are available deal with the Tamil language.

Secondly, some papers in the volume adopt macro-perspectives discussing macro-topics like “History of Language Studies” (Fernandes, Muru), and micro-perspectives discussing micro-topics like the nominal and verbal morphology of Tamil (Chevillard, James), looking at how early descriptors interpreted and codified the linguistic structures of Tamil (Chevillard, Ciotti, James), with an emphasis on the interface between Indigenous and Western grammars (Ciotti, James, Muru).

This Introduction aims, firstly, to offer a wider contextualization for the papers collected in this volume and for the topics each paper discusses. However, it is not restricted to the presentation of the five papers included here which are discussed in the last paragraph. Instead it goes beyond and also aims to offer further perspectives on the study of works by linguists avant la lettre, also posing, in this way, supplementary arguments for that change of attitudes toward early linguistic works referred by Zwartjes (2012) in his conclusive remarks. Indeed, in conclusion to his excellent survey on the historiography of missionary linguistics Zwartjes (2012: 213) states that: “the missionary linguists’ contribution to the study of language is acknowledged today, but attitudes still need to change further, and there is still much room left for further research.” For this reason, these premises are not only introductory to the volume, but they are, at a certain extent, part of it.

Hence, leaving aside this Foreword and the last paragraph in which papers are discussed, the Introduction is structured into four paragraphs. The first one introduces the topic, only briefly discussing how the diffusion of Christianity in India led to the description of South Asian languages.4 In fact, the first paragraph mainly echoes the spreading of the new linguistic and cultural knowledge derived from the production and circulation of these pre-modern texts, a circulation which occurred beyond the real intentions or awareness of the authors who composed the texts representative of it. Thus, it discusses how these descriptions blow-out into different communities and became part of a more general discourse related to a shared meta-knowledge about South Asian languages and customs, which included both the Indian grammatical traditions, the Western and also the missionary grammatical tradition.

The second paragraph emphasises the role played both by Portuguese individuals, mainly missionaries, and by the Portuguese language as metalanguage for grammaticisation. It provides a snapshot description of the diffusion of the Portuguese language in India as the lingua franca, discussing the environments in which different varieties of Portuguese were used, with particular reference to the religious and educational context.

The third paragraph reflects on the grammaticisation of South Asian languages, thus on the grammatical model of reference early descriptors used as well as on the expansion this model underwent in order to respond to the needs of describing new linguistic structures, typologically distant from early descriptors’ language of reference, either mother tongue, or the language of Western grammatical treatises. As anticipated above, the last paragraph introduces the papers included in this volume.

I am very grateful to all those who participated in the panel – both presenters and audience, as well as organisers who enthusiastically accepted it –, and particularly to the authors who submitted their articles for publication to whom I am also indebted for their significant suggestions and comments on earlier drafts of this Introduction.5

I am of course also indebted to the editors of the Journal of Portuguese Linguistics for the opportunity to publish this special issue and to the two reviewers whose remarks and comments greatly benefitted both myself and this Introduction. I would like to point out that any imperfections and mistakes are entirely my responsibility.

2. Background

“When Europeans started on their far-reaching exploratory expeditions … Their interests in the people they met comprised their religion, social organization, morals and, above all, natural resources and economic status. From the very first moment, however, language turned out to be a practical problem of great significance [italics mine]. To establish contact with the people they met and to obtain maximal benefit from the encounter, a certain amount of communication was necessary” (Hovdhaugen 2000: 925).

The situation described here for the early Travellers and Explorers6 by Hovdhaugen can also be applied to those missionaries who were the first to understand the crucial role of learning the local languages in order to spread the Gospel.

According to historical and religious sources, Christianity reached India with the apostle St. Thomas in 52 A.D. By the end of the thirteenth century the first Franciscan and Dominican friars followed him and zealously continued the work of evangelising the Indian Subcontinent.7 The first Observant Franciscans’ monastery was founded in 1518 in Goa; the city was officially proclaimed a diocese in 1539 (Zwartjes 2011: 25). Despite Dominicans’ and Franciscans’ work, the “area of Christian influence began to grow significantly only after the beginning of the 16th century, as Portuguese and Spanish seafarers extended European commerce and culture to include isolated portions of West Africa, the West Indies, and portions of South and Southeast Asia” (Gray 2000: 930).8 After the Portuguese arrival the Jesuits, along with the Franciscans and Dominicans, worked under the Portuguese Padroado Real. The Society of Jesus started building mission stations in the fortified Portuguese settlements when they reached the Indian subcontinent; however, it was the arrival of the Jesuit Francis Xavier (1506–1552) in Goa that inaugurated a new era of evangelisation in 1542. Indeed, Xavier can be considered as the first example of an illuminated missionary. He very quickly understood how crucial the learning of local languages was to the spreading of Gospel among Indians. He was the first one to apply that method for which Jesuits are so well known: the accomodatio.9 For example, he used to go around praying using a little bell which emitted a sound similar to those of bells used by Brahmins during pujas (Cuturi 2004: 7–60; Muru 2010: 30–33).

On the occasion of the Council of Missionaries in Chorão in 1575, Xavier and Alessandro Valignano (1539–1606), an Italian Jesuit appointed as Visitor of Missions in the Indies, designated Henrique Henriques SJ (1520–1600) to translate the books of the Christian tradition, such as a Catechism and Manual for Confession into Tamil. Henriques had already written a grammar10 and a dictionary of the Malabar language on Xavier’s request.11 This means that the main missionaries’ tools had already been in circulation since the beginning of the 16th century: there were grammars and dictionaries for the instruction of other European missionaries and Catechisms and Manuals for Confession, all with the aim of instructing Christianised Indians12 and spreading the Gospel in India. Behind the diffusion of these books was the missionaries’ clear intent to reduce their dependence on local interpreters, called topaz,13 and to become more effective in the conversion of the local population. Furthermore, an important innovation brought by the Jesuits was the printing press which favoured the prestige and the circulation of these tools. Indeed, in 1556 the first printing press was set up in the College of St Paul in Goa.14 Nevertheless the press was located on the West coast where other languages like Konkani were spoken with a grammar being printed in 1640,15 the first non-European language to appear printed in Indian script was Tamil.16 According to Blackburn (2003: 31), this was determined by the fact that St Francis Xavier, as well as the earliest missionaries like Henriques or Balthasar da Costa S. J. (ca. 1610–1673), mainly worked within the community of Parava fishermen on the West coast (the Coromandel coast).17 Thus, although the earliest translation of a Portuguese religious book into an Indian language was the Doctrina Cristam written in 1573 in Konkani, the first printed book in Goa in 1577 in a non-European language was in Tamil. This was Henriques’ translation of a Portuguese Catechism: Doctrina Christam em Lingua Malauar Tamul. Tambirān vaṇakkam.18 A second edition was printed in Quilon in 1578,19 followed by a further Catechism, kiricittiāni vaṇakkam (Cochin, 1579),20 a Manual for Confession (Kompessionaiyru, Cochin, 1580)21 and Lives of Saints (Flos Sanctorum, 1586).22 Also the first bilingual Tamil-Portuguese dictionary compiled by the Portuguese Antão de Proença SJ (1625–1666) was printed in Ambalacatta (Ambazhakad) in 1679.23 Descriptions of other Indian languages were printed later. For example, the first book in Sinhala appeared only in 1737, in Telugu in 1746 in Halle, Germany; in Bengali in 1778 in Hugli, and in Malayalam in 1799 in Bombay (Blackburn 2003: 34; fn. 20), while printed books in Marathi, Persian and Urdu appeared at the beginning of the 19th century. The early missionary Sanskrit grammars by Heinrich Roth (1620–1668) and Jean-François Pons (1698–1752?) (Van Hal 2016: 99) had the same fortune and were not printed before the 20th century. The only exception was the Grammatica Grandonica by John Ernst Hanxleden (1681–1732),24 which was published in Paulinus a Sancto Bartholomaeo’s O.Carm. (1794–1806) in 1790 (Van Hal 2016: 99).25 However, a conspicuous number of manuscript dictionaries and grammars composed and copied from the 16th century onwards remained unprinted: the Vocabulario lusitano tamulico e chingalatico by Jacome Gonçalves (1676–1742); the Telugu wordlists composed by Jesuits (Colas 2012a; 2012b); the Breve compendio da Grammatica Bengala, in Vocabulario em Idioma Bengalla e Portuguez (1743) written by Manoel da Assumpçam, and the Gramatica indostana a mais vulgar que se pratica no Imperio do gram Mogol (see Zwartjes 2011), along with further wordlists of Bengali and Hindustani which appeared in the second half of the 18th century composed by the British.26

As seen, the first South Asian languages to be described were Tamil spoken on the Coromandel coast and Malayalam spoken on the Malabar coast. Among the early descriptors there were travellers but mainly Jesuit missionaries whose inheritance would have been particularly incisive in Europe as shown, for example, by the quotation about the Jesuits’ early linguistic works in Hervas y Panduro’s comparative study (Fuertes Gutiérrez 2004). However, other orders too followed in the path of the Jesuits and, to some extent, of the merchants who had previously paved the way (Blackburn 2003: 27). The languages, cultural practices, religious matters, and social organisations of the Indian society were carefully recorded (or re-copied) with the purpose of providing an extensive and comprehensive description of the societies they had encountered.

Successively to the missionaries, civil servants and members of the trade European companies in India like the East Indian Company (EIC) wrote grammars of the Indian languages mainly for administrative and diplomatic purposes, describing these languages with the aim to educate both Europeans to the Indian languages and vice versa (see Muru 2018a; 2018c). Sometimes gathering the richness produced by missionaries prior to them, sometimes individually, these civil servants, as well as other missionaries like the Protestants, described the Indian languages, giving rise to new forms of descriptions and linguistic knowledge, largely influencing the macro-sociolinguistic scenario of Indian languages as well. As Annamalai (2011: 19) states “as a result of the arrival of British and other Europeans, the intellectual horizon widened in the nineteenth century and new experiences, activities and institutions were to be expressed in the native languages”. In Muru (2018c) I have demonstrated how changes in the community of addresses and purposes is also displayed in paratexts like the title page, the address to the reader, and the preface which reveal the renewed excitement and interest in languages and cultures other than those beyond the borders of actual Europe inaugurating the epoch of philological study of the non-European languages.

As already stated above, whereas many of the protagonists of the early elaboration and spreading of knowledge about South Asian languages, whose activity had started in the early 16th century and had lasted longer than the Portuguese dominion in India, were all related to the Portuguese milieu in different ways, others who got involved in this cultural enterprise belonged to different social and cultural backgrounds, like members of the EIC or the VOC. However, they directly or indirectly entered in contact with earlier Portuguese works and sometimes with the Portuguese language too. Hence, varieties of networks which acted in different places and periods which were, to a certain extent intertwined, guaranteed a temporal and cultural continuity in the production of linguistic knowledge as well as the circulation of knowledge27 related to South Asian languages and customs. Documents like letters sent to the Court of Directors by members of the EIC (Muru 2018a), as well as paratexts of second level on manuscripts and books like handwritten marginal notes on protective leaves (Muru 2018c), are evidence of this continuity in the transmission of knowledge over the centuries, a transmission which occurred behind the intention of original composers of these early descriptions.28

For this reason, I contend that detecting and describing discrete time frames and individual authorship would be quite an unproductive exercise. Furthermore, definition of authorship is a relative concept particularly within the missionary tradition as well as in the Indian context. In the latter, because most of the literature concerning language was passed on through orality and when written, in reality it was intended as rewritten, as it had always existed. In the Western missionary tradition because the only recognized authorship was the Company of Jesus.29 This could be the reason for which, with the exception of some peculiar works recognized within the missionary community as representative of a grammatical or lexicographical tradition, most of the Jesuits works were not signed or marked with authorship.30 Apart from Tamil-Portuguese or Portuguese-Tamil dictionaries available nowadays in Goa at the State Central Library or in Paris, other examples are represented by Arte Canarina na lingoa do Norte (17th cent.) composed by a Franciscan or a Jesuit, and first printed in Nova Goa in 1858, or the Grammatica ou Observações Grammaticaes sobre a lingua de Concana, mentioned in Cunha Rivara (1958[1858], see Fernandes this volume) maybe by the Italian Carmelite Francisco Xavier (or Xaver) (Zwartjes 2011: 47), and the Gramatica indostana a mais vulgar que se practica no Imperio do gram Mogol offerecida aos muitos reverendos Padres Missionarios do ditto Imperio (1778a) or the Gramatica marastta a mais vulgar que se practica nos reinos do Nizamaxà, e Idalxà (1778b).31 Only from the 18th century onward, and with the advent of printing, did the authorship assume other values and the attention and focus on the ‘agent’ of the contents of the printed book increase (Muru 2018c). Furthermore, even where authors gathered previous sources for their work, they rarely mentioned or referred to them. An example is Domingo de Madeyra SJ (1685–175?) who only refers to Giuseppe Costanzo Beschi SJ (1680–1742) in his prologue to the reader as a representative of a deep knowledge of the Tamil language, leading the reader to believe that he composed the dictionary, while most probably the Vocabulario thamulico lusitano para uzo dos missionaries da Companhia de Iesu. Composto, e augmentato pello P. Domingos de Madeyra da mesma companhia (1750) should be considered a version of Proença’s dictionary (see James 2000: 105).

For these reasons, I not only preferred to consider each one of these works as a further contribution to what had been done before, sometimes enriching what had been previously, but also, I propose to look at these sources with a methodological approach which focuses on continuity and loans in production and circulation of knowledge across diverse geographical areas and times.

Considering these assumptions, the production of grammatical and lexicographical works should be considered as part of the same discursive community as I have already stressed in Muru (2018c), especially among Jesuits, as those works were not only widely circulated, but they also presented common structural features, following which new networks for the circulation and production of knowledge were created and the previous works were englobed. I am not sustaining that the individuals involved in these networks and in this production consciously worked with the aim of creating a network, this happened as a consequence of the circulation of these works. As already pointed out above, it remains a well attested fact that early works written by some individuals were written, re-written, copied, edited, enlarged and continued by others in later periods and also in other geographical areas. Apart from Aguilar’s example stated in footnote 28, another one is represented by Antão de Proença’s dictionary (1679) which was copied several times, at different periods and in various areas. Although most of the time the copyist’s name is missing, the actual collocation of manuscripts and their provenance allow us to imagine the large trajectories these copies had as well as the extension of the networks within which they circulated.32

3. Portuguese as metalanguage of grammaticisation

When dealing with early descriptors and descriptions of South Asian languages, another issue which should be taken into consideration is the crucial role played by the Portuguese language33 as metalanguage in this process of grammaticisation. As also stated by Assunção and Fernandes (2017: 61) the Portuguese language had a determinant role in the description of many languages spoken in the East, a fact which is not comparable to any other European language. Despite Pytlowany and Van Hal (2016) having recently demonstrated how the Dutch34 were also important in this respect, as well as Annamalai (2011) who highlighted how the English language as well as the British were determinant in the Tamil Renaissance, the Portuguese language remains the most used in these early works. The main reason is not only determined by the fact that the Portuguese were the first ones to describe these languages, but also by the fact that Portuguese functioned as a lingua franca. Its role was crucial in the Indian landscape as a language of exchange and communication, as the many Portuguese-based Creoles demonstrate.35 Furthermore, as Ansaldo and Cardoso underline (2009: 6) “it is too simplistic to assume that one single variety of Portuguese would have been exported from Portugal all the way to the Far East”. As reported in Pfänder and Costa (2006: 1156), when discussing the Portuguese used in the Protestant mission on Tranquebar, this was far from being homogeneous. Indeed, the Portuguese language in India was represented by different varieties strictly correlated to the social networks within which they developed and were employed as one can assume from Baxter (1996: 301, quoted in Cardoso 2014: 108):

“[…] it seems highly likely that continua of L2 Portuguese arose, ranging from rudimentary L2 to L2 approximating to L1 European Portuguese, according to the degree and nature of contact. It also seems possible that the lower levels of the continua may have gelled (leading to pidgin formation). Varieties of L2 Portuguese would have provided significant input to L1 acquisition for certain groups of the local population. Where the L2 (and L1) input was significantly watered down in comparison with the original Portuguese L1 model of the colonizers, L1 acquisition by locals led to creolization. Continua of local L1 varieties of ‘Portuguese’ would have been formed.”

Tomás (2009) and Clements (2009) also have exhaustively described the network within which these varieties were employed, giving evidence of how the development of Portuguese-based Creoles in Asia, but primarily their linguistic similarities, can be explained throughout the network model, rather than the monogenetic theory. Like Silva Jayasuriya (2008: 172), Tomás (2009: 26) defines the Portuguese dominion as “an Empire which, in its essence, was based on a network system rather than on the control of a territory for the production of goods”. This allowed the spreading of the language, carried by single individuals who belonged or entered into different communities to spread it. Indeed, since the 16th century, Portuguese had served as the lingua franca in India within various environments, which can be identified by four main representative ones in which the functional role played by the Portuguese language is evident.36

In trade and domestic environments, the variety of Portuguese used was the one described as Portuguese Jargon – as a lingua franca (Cardoso 2014: 93). In the commercial environment, multiple and intensive exchanges in oral and written forms took place between the Portuguese and Indians, and later between British merchants of the EIC with Indians as interpreters. Indeed, even after the decline of the Portuguese domain, the British continued to use the Portuguese language, until at least the second half of the 17th century (Cardoso 2016: 71, fn. 1; Muru 2018a).

Private life was another environment where Portuguese was commonly used, especially in its oral form. Indeed, it was the language used within inter-racial marriages/unions. These were the environments in which, much more than in others, pidginised varieties were used and where Portuguese Creoles arose. In Cardoso (2008: 46), the author provides a list of these Portuguese-based Creoles, reminding us how “the combined efforts of Hugo Schuchardt and Monsehor Sebastão Rodolfo Dalgado meant that now have late 19th–c. and early 20th–c. data for several of these varieties”. Within this environment the role women played was determinant as Tomás (2009) demonstrates. Indeed, the nuptial relationships were the more efficient ways through which the Creoles developed and diffused, a practice also common among the Dutch.

Another environment where Portuguese was employed was education, mainly delivered by missionaries. Despite its formal context, a pidginised variety of Portuguese was used. British civil servants and priests, and the Protestant Dutch referred to this variety as patois (Nurullah & Naik 1951 [1943], 1964 [1945], Penny 1904). The Portuguese patois was a ‘corrupted’ or ‘bastard Portuguese’ (Cardoso 2008: 45) quite different from that spoken in Portugal. According to Penny (1904) this was the medium of instruction in Jesuit Portuguese Colleges between the 16th and 17th centuries in the Madras Presidency (on the East Coast, see Muru 2017, 2018a), but also the language used in Dutch and British schools.37 Most recent studies, like Cardoso’s (2014: 94), have argued that the variety used in school and in religious contexts “was especially acrolectal in the Malabar, the southwestern coast of India”. The acrolectal variety was also used in writing: in religious texts, in commercial letters, and in private correspondence (Cardoso 2014: 91).

Undoubtedly, the study of Portuguese was prioritised, above all in South India, for those Europeans arriving there. Two examples are found in the Protestants Bartholomæus Ziegenbalg (1682–1719) and Christoph Theodosius Walther (1699–1741)38 who worked as missionaries in Tranquebar for the Dutch mission.39

Bartholomæus Ziegenbalg wrote a few weeks after having arrived in India in 1706:

“When we […] reached land […] and the Malabar heathens saw us, there was much discussion among them about who we were and why we had come here. In the beginning we couldn’t talk to them at all because we only understood Danish, but not Portuguese or Malabarian. We therefore made it our first concern to learn the former as quickly as possible.” (quoted in Pfänder & Costa 2006: 1157).

While Walther asserted that he had to learn even four languages: two varieties of Portuguese, the high one represented by the ‘European’ Portuguese and the low one corresponding to the basic variety used in daily conversation. It was the same with the Tamil language of which he had to learn both the high and the low variety (Van Hal 2016).

Consequently, as it occurred for Jesuits who, despite their provenance, used the Portuguese language as metalanguage for their descriptions,40 other missionaries like the Protestants used and learnt the Portuguese language because it was the main tool through which one could gain access to the previous linguistic sources. Indeed, Portuguese was also the language of the grammaticisation of many Indian languages. As Županov and Xavier Barreto (2015: 205) observe:

“By the time the British arrived and settled down on the shores of the Indian Peninsula, before enthroning their own linguistics order missionaries working under Portuguese royal patronage and, from the late seventeenth century, those who were sent by the Roman Propaganda Fide had already mapped a dozen or more of Indian and Asian languages with the intention of ‘converting’ them”.

The reason this is important relies on the fact that Portuguese, being used in the grammaticisation of many South Asian languages, was also the metalanguage used for describing those linguistic features typologically distant from the Indo-European languages and from which new labels for referring to these new linguistic features or categories were created (see Chevillard this volume).

Further evidence that one has nowadays of the ‘primacy’ the Portuguese language held in India is also provided not only by the testimonies left by missionaries and civil servants on the need to learn this language, but also by the written documents, mainly commercial ones, held in different archives in India (Ernakulam)41 or in England (Kew Archive or British Library). Most of the correspondence, despite having been written between the 17th and the 18th centuries and despite being representative of transactions occurring between the EIC or the VOC and Indians (both merchants and Princes), was still written in the Portuguese language (Muru 2018a).

This primacy of the Portuguese language remained unquestioned for centuries at least until the beginning of the 18th century when the British set up their first English schools and the British commercial domination of India progressively turned into a territorial/political one with the Battle of Plassey (June 23rd, 1757) (Muru 2018a). As discussed above, this means that the Portuguese language survived the Portuguese dominion, obliging other Europeans who reached India to learn it and to employ Portuguese speakers during their voyages as translators (see Lopes 1969 [1936]).

With the increasing spread of the English language (see Muru 2009, 2018a) the affirmation of its roles as lingua franca, as well as the spreading of other European languages like French and Dutch, in territorial, political and linguistic settings important changes in the history of knowledge transmission occurred: not only was the Portuguese language replaced by the English one (or to a lesser extent, other European languages like French and Dutch),42 but grammars ceased being mostly anonymous and written in Portuguese. As such, the Western Grammatical Tradition (hereafter WGT) was increasingly integrated with the indigenous Indian Grammatical Traditions (hereafter IGT).43 While this practice had previously been initiated by Jesuit missionaries like Giuseppe Costanzo Beschi for Tamil, it increased even further after the 18th century.

4. The model of reference in the grammaticisation of Indian languages

In the composition of these early grammars, Portuguese was the predominant metalanguage which was largely but not exclusively used.44 For example, Beschi, Hanxleden,45 Ziegenbalg, and Walther used Latin, while, as stated above, other metalanguages like French, English, and Dutch also appeared, even though these works had less circulation. I have already discussed the reason why the Latin language was used instead of the Portuguese one in Muru (2018c), so here it is enough to say that this was the language of the main grammatical framework that most Jesuit missionaries considered in the composition of their grammars.

As Law (2003: 231) points out, during the Renaissance the rediscovery of Latin had increased the awareness that “it had not always been the refined and flexible tool, ideally suited to rhetorical expression and logical argumentation alike, that it was in their own day”. This meant that, like the vernaculars, the Latin language had been a ‘rude’ language which had reached a certain status and perfection only through grammaticisation. In other words, Latin increasingly gained formal perfection through linguistic activity, which progressively led to the production of linguistic tools (Auroux 1994: 9). The increased awareness that Latin had not always been ‘a refined and flexible tool’, along with the “switch of interest from the universal to the particular, from Language to languages, took place simultaneously with the discovery of new worlds across the oceans and of the full extent of linguistic diversity beyond Europe” (Bossong 2007: 125). As had already happened in Europe where the vernaculars were grammaticised according to the Latin grammatical framework, similarly missionaries strived to grammaticise those non-European languages they encountered through linguistic tools elaborated from other languages (see James and Muru this volume). As such, grammatical concepts, technical terms, and the organisational structure were transferred from one language description to another. The first missionaries were mainly guided by pragmatic and communicative needs and were generally not moved by theoretical aims when compiling their grammars and describing other languages.46

However, applying the Latin or Vernacular model of reference, they were forced to modify and extend it in order to give an account of those typological features of the new languages that could not be found in Latin or Portuguese. Nevertheless, the application of this common model of reference led to a remarkable homogeneity in the description of several typologically different languages around the world, the linguistic diversity of those languages urged the missionaries to embark on a deeper theoretical reflection regarding the description process. New categories needed to be introduced in order to create new linguistic tools which could more aptly account for the diverse structures and sounds (phonemic idiosyncrasies) of non-European languages.

The discovery of Babel, primarily by missionaries (Firth 1937), threw the universalist version of the Modists into crisis. They saw in Latin, as described and analysed by Priscian, a universal infrastructure able to explaining how all languages communicate in the same way, allowing, independently from their diversity, the abstractions within the mind (modi intelligendi) of the modi essendi of things which are communicated through the modi significandi of the language (Robins 1971 [1961]: 119). Consequently, when the number of empirical data increased, it became difficult to sustain the previous interpretation of the Babel history in which all the world’s languages were created through confusion and preserved few elements in common, demonstrating the origin of a common language, the lingua Adamica. The diversity became evident through the non-European languages which could not easily fit into the universal Latin model, leading missionaries and later descriptors to expand the linguistic terminology and perspective (see Ciotti this volume).

The Renaissance, during which missionaries described the newly discovered non-European languages, was a breaking point in the idea of a universal language developed during the Middle Ages.47 Indeed, during this time, the idea of the previous philosophical behaviour towards the language that had characterised the Middle Ages changed, and even the divine origin of the language of the Dark Ages was gradually overtaken. Following the explorations of the 16th century and during the Renaissance “the language was assumed to be a philological object, rather than the medium of the logic. It was conceived as a social-historical institution rather than the medium of knowledge and expression of thoughts. […] For this reason, the attention moved from the language as a universal/general expression, toward languages and their multiple historical forms” (Coseriu 2010 [2003]: 198–199). As a consequence, after the explorations initiated in the 16th century, the universality of language, as it was previously conceived, was questioned due to the fact that it was represented by a variety of category systems, and the presence of diverse abstracts rules were ultimately acknowledged as the root of all language.

Although the Latin grammar continued to represent the backbone of each language description, more often than not missionaries and early descriptors contributed to the development of new grammatical concepts or technical terms through their understanding and recording of linguistic peculiarities. In the Indian context in particular, the process of grammaticisation and the transfer and extension of linguistic tools happened in a bi-directional way. Indeed, as Zwartjes (2011: 27) states “in contrast to the practice in the Americas, in Asia the Portuguese missionaries could benefit from local grammatical traditions” and they gradually conveyed indigenous grammatical terms and integrated them into the Latin/vernacular grammatical terminology.

However, in order to fully appreciate and understand not only the contribution of earlier descriptors of South Asian languages within this perspective, but also how the linguistic descriptions and thus the Science of Language changed throughout the centuries by the application of grammatical models originally elaborated for other languages, it is necessary to be aware of which model of reference was applied. As discussed in Muru (2018b) this can be a challenging task when looking into early descriptions of South Asian languages by Westerners, since the authors of these grammars rarely mentioned their sources. Thus, the challenge is to understand which of the many Latin grammars available in Europe between the 16th and 17th century was used as a model of reference. An insightful examination of these Latin grammars would be complicated but desirable, since it might provide a sound answer to the challenging questions regarding the models of reference in use at the time of the descriptive enterprise.

As pointed out above, even though direct references to other works in the early descriptions, like, for example, in the Malyalam manuscript grammar kept at the State Central Library in Goa and in the Gramatica Linguæ malavarice ʃamʃcredam by [Jo]giuʃeppe Chariati Indiano (17th cent.)48 are unlikely to be found, there are some intriguing exceptions at least in Tamil works. Examples are those represented by the works of Henrique Henriques (16th cent.) and Balthasar Da Costa,49 while most of the time, authors refer to previous works in a very generic way, as Proença does when he refers to the source used in the compilation of his dictionary in his address to the reader.50 Henrique Henriques’ work specifically suggests to his readers that they have to be acquainted with Grammatica da lingua portuguesa (1540) by João de Barros,51 referring in this way to the WGT. At the beginning of his Arte he writes:

Para mais facilmente se emtender esta arte ha mister ter conhicimento da arte Latina, e os que nõ souberẽ Latim devẽ de leer por a gramatica portuguesa feita por Yoaõ dBairros. [MS COD. 3141, fol. 6v].

[To understand this Arte more easily one should have knowledge of the rudiments of Latin. Those who do not know Latin should read through the Portuguese grammar composed by Jõão de Barros].52

Differently from Henriques, Costa and Proença’s works provide two appealing examples of the practice of cross-referencing other missionaries’ work, more specifically in Costa’s and in a more generic way in Proença’s. Rather than referring to well established Latin or Portuguese grammarians, they address what I defined as the Missionary Grammatical Tradition (MGT) in Muru (2018b; 2018c). Indeed, Costa makes clear mention of Gaspar de Aguilar’s Tamil Arte (17th cent.),53 while Proença openly refers to several missionary works as partially representative of his sources in the compilation of the dictionary.

“Prometo mais copia nas palavras, por que alẽ de por todas, as que em outros vocabularios achey, principalmente no que compôs com incesãte trabalho, no discurso de mais de trinta annos, o P. Ignacio Bruno (que depois governou duas veses esta Provincia) […] pús todas as, que há nas obras do P. Roberto Nobili & do P. Manoel Martins livros por onde, os que nesta missam andamos, aprendemos o melhor fallar. Li pera este effeito muitos livros dos naturaes, finalmẽte acreçentey, os que no exercisio do fallar, e tartar com os naturaes alcancey, e vou alcancando.” (MS. Borg. Ind. 12, fol. 5v)

[I promise to list a greater number of words, because of additions to those found in other vocabularies, principally in the one composed with tireless endeavor in the course of more than thirty years by Fr. Ignacio Bruno (who later twice governed this Province). […] I have entered all those found in the works of Fr. Robert Nobili and fr. Manuel Martins, books through which we, who work in this mission, learn the best speech. For this purpose, I have also read many books of the natives; finally I have added those words which I have acquired or am acquiring by talking and dealing with the natives].54

Lack of references to previous works diminishes after Beschi’s works since grammarians always refer to him as the representative of grammatical and lexicographical tradition, whereas the pre-Beschi works are only referred to in a generic way.55

Despite this lack of references, it is generally accepted by the academic community that the abridged version of Manuel Álvares’ De Institutione grammaticae represented the main model of reference among missionaries within the WGT. This must have been the case, at least since the second half of the 16th century, when Álvares’ De Institutione was codified by the Constitutiones (1546 and 1552), Ordo studiorum (1565) and the Ratio Studiorum (1586, 1591, 1599).56 From then, his works became the main pedagogical manual that was used in Jesuit colleges; its wider diffusion and translation into several languages throughout Europe followed.

Zwartjes (2002: 28–29) suggests that at least some missionaries were generally familiar with the works of Donatus, Priscian, and Quintilian, as well as being aware of the existence of Valla’s, Nebrija’s, Despauterius’, and Erasmus’ works. However, it is the abridged version of Álvares’ De Institutione known as Ars minor (Kemmler 2015), that presumably provided a robust model of reference for both Portuguese and Spanish missionaries working on the description of non-European languages. For this reason, the Alvaristic model and its extension, should be taken into consideration while looking at the early descriptions of South Asian languages from the 17th century onwards.

However, the IGT should also be borne in mind. Indeed, within the Indian context, missionaries entered in contact with Tamil and Sanskrit tradition (IGT) initially through the medium of orality and Indian teachers (mestre) to whom some of them refer (e.g. Proença in his address to the reader). Hence, early descriptors became gradually aware of the IGT when, from the late 17th century onwards, they acceded the primary Indian sources in person. Evidence of this interface between the WGT and IGT is found in the works by the German Jesuits Roth and Hanxleden, as well as in the works by the Italian Jesuit Beschi.

Vielle and Van Hal (2013) point out that “as stated himself at the outset of his work, Hanxleden relied on the Sanskrit grammar Siddharūpa, of which he had copied and annotated a manuscript” and which was the grammar used in Kerala for learning Sanskrit. Roth also showed perfect familiarity with the technical terms of the Sanskrit grammatical tradition (see Zwartjes 2011: 27), along with Aguilar/Baldaues who described the Tamil case system using all three grammatical traditions: the Latin, Tamil, and Sanskrit ones (see Muru 2014b: 371–375).

Whereas the Italian Beschi consistently took part in the production of knowledge on the Tamil language from the early 18th century. Indeed, he studied the Tamil language for more than twenty years and devoted all his life to writing grammars, bi-/trilingual dictionaries, poems and religious-moral treaties in Tamil.57 As highlighted in Chevillard’s contribution (1992a), an example of the dialogue between IGT and WGT emerges in Beschi in the treatment of the appellative verb. Furthermore, although the previous missionaries of the Tamil language such as Henriques, Costa, Aguilar, and Proença (Muru 2010, 2014b) had already acknowledged the presence of different registers of the Tamil language,58 it was Beschi59 who was the first to provide two official different grammars for at least two varieties of Tamil: a grammar of koṭunTamil, meaning of the lower register of Tamil, and a grammar of cenTamil, meaning a grammar of the higher variety of Tamil. 60

The intersection between these grammatical models, the WGT, the MGT, and IGT, had an impact on the Indian linguistic studies. For example, applying the Latin model to the description of Tamil allowed missionaries to identify linguistic categories of the language that would have also entered within the later grammatical tradition of Tamil (see Muru in this volume). Evidence of this impact is represented by the renovation, for example, of the writing system. Before the arrival of the missionaries, the Tamil script did not differentiate between long and short e/o, apart from the earliest attempt in codifying this distinction in writing (e.g. Proença’s diacritics used for differentiating C+[ē, ō] and C+[a] vs. CC or C, Muru 2010: 62; 194–195), but with Beschi the reform of the writing system definitely occurred. After his input, e/ē and o/ō was to be graphically differentiated into / and / respectively.

As conclusive remarks, I want to summarise the three main points the contribution that early linguistic works, starting with missionaries, have made to the Science of Language.

The first one refers to all that has been described in the previous paragraph and relates to the bi-directional and productive exchange occurred between European and Indian traditions which had as consequence a reciprocal influence.

The second one relates to the sheer amount of linguistic data which, thanks to the linguistic fieldwork carried out by missionaries before and others later, reached Europe, enriching the knowledge about language as a human faculty, and thus contributing to the development of new ideas about language and its origin, from a divine one whose diversity had been determined by the confusion of the Tower of Babel, to a socio-historical institution.

The third one, as direct consequence of the previous one, relies on the contribution given to the understanding of how languages and thus human language as a human faculty work that occurred also to the larger amounts of diverse linguistic data. Indeed, the breadth of the reflections about linguistic features which were typologically divergent from Latin or European languages undoubtedly contributed to the development of typological studies. The contrastive approach adopted in the descriptions of non-European languages,61 typologically divergent from their model of reference, allowed the detection of many linguistic peculiarities which had remained unknown until then. This need also led to the elaboration of new linguistic terms more suitable for the description of the linguistic diversity which are still used nowadays, as was the case for the term postposition discussed in Zwartjes (2002).62

5. Papers in this volume: aims and contribution

The papers in this volume, focusing on the early descriptions of South Asian languages, present the results of recent research in which authors attempt to shed light on various aspects related to early linguistic works of non-European languages. As pointed out in the Foreword, the papers focus on different languages, three deal with the Tamil language (Chevillard, James, Muru), and two with Indo-Aryan languages, Sanskrit and Konkani respectively (Ciotti, Fernandes) and their main contribution is represented by the different angles through which they outline and debate the analysis of early linguistic works of South Asian languages. The following perspectives are discussed in this volume:

  • the application of the Latin grammatical framework (Chevillard, Fernandes, James, Muru);
  • the strategies and innovation applied by early descriptors in the description of non-European languages (Chevillard, James, Muru);
  • the contribution and impact of early description in later grammatical works or on the Indigenous linguistic tradition (Ciotti, James, Muru);
  • the interface between Western and the Indigenous grammatical traditions (James, Muru, Ciotti);
  • the dissemination of new linguistic knowledge acquired through collectors of early descriptive works (Fernandes, James).

Fernandes applies a macro-perspective and contributes to the History of Language Studies by meticulously following Joaquim Heliodoro da Cunha Rivara’s (1809–1879) life, work, and activities for the promotion of the Konkani language and the diffusion of knowledge related to it. He underlines the crucial role the Portuguese physician, professor, librarian, and politician Cunha Rivara played in the promotion of the status of the Konkani language which he carried out throughout the editing and diffusion of earlier works on Konkani, three grammars and one Konkani dictionary, among which the Arte da lingoa Canarim (1640) composed by Thomas Stephens SJ (1549–1619).

The editing and circulation of this work, as well as the strenuous activity of the collector Cunha Rivara, contributed to the creation of something which went beyond Cunha Rivara’s intentions, that is a renovation of previous sources and a network of the transmission and diffusion of linguistic knowledge on Konkani. Despite the macro-perspective adopted, in discussing the earlier works on Konkani that Cunha Rivara collected, Fernandes also provided insightful details about how and to what extent these early grammars relied on the Latin grammatical framework for the description of Konkani.

Compared to Fernandes’ contribution, the other papers adopt a micro-perspective since they focus on internal issues of the descriptions of South Asian languages. They intend to provide a comprehensive picture of the descriptive issues that these linguists avant la lettre had to face: notably the difficulties early descriptors underwent when confronted with the less familiar features of languages like Tamil or Sanskrit. In this sense, James and Chevillard elucidate the creativity and innovations proposed by individual descriptors, while Muru and Ciotti stress how the observation of these less familiar linguistic features impacted on the Tamil grammatical tradition and the Western grammatical tradition respectively. Hence, James and Chevillard focus on nominal and verbal morphology of Tamil respectively, Ciotti on morphophonemics of Sanskrit, and Muru, following a discussion on the internal organisation of early grammars of Tamil, focuses on verbal morphology.

James offers an insight into the description of Tamil declension, paying particular attention to the ablative case of Tamil. Highlighting the difficulties, the early descriptors encountered, which sometimes led them to the wrong interpretations of some features, James provides an overflow of data taken from linguistic works over a span of four centuries (16th–20th cent.). In particular, stressing how early descriptors were not always able to distance themselves from their Latin model of reference, he demonstrates how some of their (mis)understandings passed on to later works. Indeed, comparing earlier descriptions with later ones and taking the Tamil grammatical tradition into account, James clearly demonstrates the line of knowledge transmission which characterised works on the Tamil language – which I referred to at the beginning of this introduction. In fact, he shows how the representations made on Tamil by the Portuguese on the basis of Latin in the mid-sixteenth century provided the foundation for all subsequent European analyses of Tamil until the twentieth century.

Similarly, Chevillard observes the difficulties early descriptors encountered when dealing with the Tamil verbal morphology, above all in the lemmatisation of verbal forms within bilingual dictionaries. But differently from James, he points out the strategies the early descriptors adopted and how they progressively mastered the complexity of the Tamil verbal morphology, making use of labels taken from Portuguese or Latin terminology. Consequently, he highlights how early descriptors applied the WGT of reference and gradually improved in the identification and categorisation of linguistic features like infinitivus, or the Tamil paired verbs which some modern linguists have referred to as affective and effective.

The Tamil paired verbs also recur in Muru’s contribution. Similarly to Chevillard, she focuses on the tools which the WGT provided for the early descriptions, demonstrating how they initially influenced the internal organisation of the early grammars and the categorisation of Tamil. However, she also points out how the early descriptors figured out categories like mood and aspect, or the differentiation between weak and strong verbs which are lacking in the Tamil grammatical tradition, not only because of the application of the Western grammatical model, but also for the missionaries’ intuitions and skills.

Ciotti moves forward and discusses later descriptors of Sanskrit like William Dwight Whitney (1827–1894) and Friedrich Max Müller (1823–1900). He offers fresh insights into the process by which grammarians elaborated the categories of internal and external sandhi, after having being described by earlier descriptors who, however, did not find a particular label for them. Ciotti shows how sandhi became gradually categorised into the WGT, with specific terms directly derived from the encounter with the IGT. He also points out how these labels had a profound impact on later important works representative of Linguistics studies like Bloomfield’s Language 1933 and Menomini Morphophonemics 1969.

If micro and macro perspectives adopted by authors in the observation and discussion of early linguistic description of South Asian languages are moved by different questions and lead to different answers, there is a common purpose that the five papers included here share. It is to demonstrate and support how the earlier descriptions and descriptors and the role played by later intermediaries can contribute to enrich the History of Language Science, but also contribute to other fields of research within a multidisciplinary approach. They provide different snapshots on the history related to the composition, elaboration, transmission, and diffusion of early linguistic works. They also offer an insight into the descriptive strategies and the internal organisation the early descriptors adopted in the effort to understand South Asian languages. For these reasons, the research included in this volume represent valuable sources, especially for the study of languages strictu sensu, where historical linguistics and sociolinguistics, but also history, and sociological studies as well, would undoubtedly benefit.

In conclusion, in the light of the various array of issues these contributions touch upon, we hope they will open up not only new margins for interdisciplinary dialogues, but also encourage more scholars to undertake the difficult task of working on early primary sources of non-European languages in order to contribute new insights to the History of the Language Science.