Transliterating Cities: The Interdiscursive Ethnohistory of a Tamil Francophonie

Abstract

The interdiscursive ethnohistory of outdoor signs and other transliterated graphic artifacts from four urban neighborhoods in Puducherry, Paris, and Montreal is based on linguistic, ethnographic, and archival analyses of disparate sociohistorical contexts in which businesses and organizations promote or devalorize printing in Tamil and Roman scripts. Signs that project the image of a Tamil francophonie depend on structures of addressivity that animate graphic artifacts and potentially lead to new encounters between francophone Tamils. Thus, transliterations into Tamil, French, or English recalibrate the chronotopes of francophone Tamil settlements. Embodying the present, Paris provides the grounds for reproducing the linguistic community through adherence to International French, despite its paucity of transliterations. Montreal’s transliterations embody the diaspora’s future, emphasizing vibrant entrepreneurial activities in grassroots literacy, whereas signs in Puducherry featuring ornamental displays of French offer opportunities to connect with a past in which Tamil and French once coexisted in colonial handbooks and streets.

Unlike some Indian languages, Tamil does not have a single standard transliteration system. Authoritative sources … use different transliterations, especially for some of the laterals and rhotics, where true confusion reigns. To make matters worse, popular transcriptions, such as those used in public signing, transliterations of person names, etc. typically do not mark differences in vowel length, retroflexion, or other distinctions. This is unfortunate, but scholars and others have not been able or willing to agree on a standard transliteration.

(Schiffman 1999, 12–13)

The three francophone cities of Puducherry, Paris, and Montreal, although divided by geography, politics, and history, share similar “linguistic landscapes” (Landry and Bourhis 1997; Gorter 2006; Shohamy and Gorter 2009), or publicly viewable signs, located in neighborhoods where Tamil-speaking communities regularly engage with francophone and anglophone communities. Featuring diverse signs composed in Tamil, French, and often English and printed in both Tamil and Roman scripts, these linguistic landscapes represent complex permutations of translations (defined as the conversion of meaning from one linguistic code to another) as well as transliterations (defined as the conversion of words from one script system to another). The signs encompass (1) French words translated into Tamil, (2) Tamil translated into French, (3) Tamil transliterated into Roman, (4) French transliterated into Tamil, (5) English transliterated into Tamil, and (6) French translated into English yet transliterated as Tamil script. Analyzing multilingual signage in the neighborhoods of these three francophone cities—the only known ones in the world to display linguistic landscapes that include both French and Tamil—supplements existing scholarship on the history of Tamil print culture (Venkatachalapathy 1994; Blackburn 2006). Among other insights, these examples reveal patterns of similarity and discontinuity suggestive of how transliteration practices are entangled with language politics at transnational or global scales.

In the opening quote, sociolinguist Harold Schiffman laments the absence of conventions for transliterating colloquial Tamil into English, highlighting the predicament of business owners who cannot print “correct” bilingual outdoor signs in Roman script due to this lack of standard. His survey of transliterations between Tamil and English in South Asia delineates key areas of disagreement in the graphic representation of initial, geminate, and postnasal stops; glides; and oral, nasal, and long vowels (Schiffman 1999, 12–20), and he attributes the nonconventionality of transliteration schemes to the phonological and morphosyntactic idiosyncrasies of colloquial Tamil, the variety spoken by most Indians (Britto 1986; Annamalai and Steever 2015). Instead, Shanmugam refers to the “problem of transliteration of Tamil … as a problem of spelling” and advocates for adopting the Tamil Lexicon system established in Madras in 1924, denouncing other systems seeking to replace diacritics with diagraphs as “alarming,” “cumbersome,” and not a “step in the right direction” (1975, 57). Even though Schiffman endorses a mostly descriptive view and Shanmugam a prescriptive one, both would agree that the standardization of Tamil transliteration is a critical step toward modernizing Tamil and advancing Tamil studies.

Despite these astute observations, neither scholar has sought to further contextualize the history of transliteration by identifying whether colonial and postcolonial tensions precluded its standardization or by documenting reoccurring patterns across South Asian and diasporic signage to discern how typographical forms might pertain to regional language policies.¹ I propose an interdiscursive ethnohistorical approach that would start comparatively by noting the fierce opposition exhibited toward the transliteration of Tamil into Devanagari in Tamil Nadu and Sinhala in Sri Lanka,² but not into Roman. This approach would then ask why the transliteration of Tamil into French and English on nongovernmental signage leaves unresolved the question of representing diacritics, Grantha letters, and sandhi. Research on Tamil diasporas or colonial French India would suggest that most transliterations to and from Tamil involve not just one language but two—in this case, French and English—and argue that one must analyze the many permutations of variation across script and typographic choices. By regarding transliteration to be a practice of cultural translation more broadly conceived (Asad 1986), I add that transliterated signs also act as windows that refract, rather than reflect, imperial struggles and colonial legacies and must be considered part of generative chains through which semiosis begets new signs through interpretants (Agha 2005a; Parmentier 1994; and see also PWP, 99). This view helps to explain the variegated impact of French politics on Tamil print culture.

The first Tamil-to-French transliteration conventions devised in the late nineteenth and early twentieth centuries in South Asia presented notable incongruencies with later French-to-Tamil transliterations, particularly in the typographic choices involving rhotics, glides, and vowels. In this article I first survey the different systems for transliterating the Tamil alphabet into Romanized French in several grammar books and manuals printed by colonial presses in Pondichéry and Paris and compare these with French-to-Tamil transliterations of toponyms on contemporary street signs found in Ville Blanche (White Town), a neighborhood in Puducherry, to assess their degree of conventionality. Although colonial era Tamil-to-French transliterations were unique in their endorsement of diacritics and combination of literary and colloquial Tamil orthographies, the influence of British phonetic conventions increasingly became the defining trademark of French-to-Tamil transliterations in the postcolonial period. Hence, I next examine the mediating role of English in French-to-Tamil transliterations, first by comparing storefront signs in two Sri Lankan Tamil neighborhoods located in Montreal—Parc-Extension and Côte-des-Neiges—taken before the end of Sri Lanka’s civil war in 2009 and again ten years after the war. Capitalizing on the changing nationalist politics in Canada, India, and Sri Lanka, businesses exhibiting more self-assured and outward-looking stances in the postwar years experiment with stylistic variation while adhering to local laws. Finally, I analyze storefront signs in La Chapelle, a Parisian neighborhood where businesses established since the 1980s by Sri Lankan Tamils and 1950s by Indian Tamils from Puducherry showcase signs that foreground printing in Roman script and deemphasize transliterations into Tamil. Although Paris is recognized by francophone Tamils for the prestige of International French and new literary voices translating or writing in both Tamil and French (Shanmuganandan 2003), the lack of a migratory pull factor there limits the extent to which Parisian streets can renew their past cultural relationship with Pondichéry, a French Indian city fading out of collective memory, or forge future social and business ties with Montreal, where francophone Tamils around the world increasingly call home. This complex spatiotemporal relationship, which involves the reconfiguration of Puducherry as a city of the past, Paris as a city of the present, and Montreal as a city of the future, can be observed in the transliterations of outdoor signs.

The Method and Theory of Interdiscursive Ethnohistory

Patterns of typographical form and script choice found in transliterations index complex and shifting interdiscursive and ethnohistorical entanglements of empire, nation, and diaspora. Deconstructing these links and entanglements requires a new type of semiotic inquiry, which I refer to as “interdiscursive ethnohistory.” Informed by linguistic, archival, and ethnographic methods, this conceptual framework acknowledges that the social life of “indexical orders” (Silverstein 2003) is always potentially global. The interdiscursive ethnohistory of signage in Tamil, French, and English thus situates the analysis of print media in and across disparate sociohistorical settings of language contact to ascertain whether the interdiscursive relationships between entextualized signs produce far-reaching networks of history and genealogy (Briggs and Bauman 1992). Similar to studies of debates over orthography (Schieffelin and Doucet 1994; Jaffe 1996; Romaine 2002; Jaffe et al. 2012) and script choice (Fenigsen 1999; Daveluy and Ferguson 2009; Choksi 2015) and “fontroversies” over typeface (Murphy 2017), there are two approaches that one can take with this. A language ideological approach analyzes transliterations as strategic choices that mediate the rivalries associated with different colonial and postcolonial printing regimes informing the visual representation of codes and scripts. By instead regarding the material aspects of printing as sociosemiotic phenomena whose aesthetic qualities fall under the Peircean typology of icon, index, and symbol (Murphy 2017), one can otherwise analyze how certain signs connect spaces and beckon toward new, existing, or forgotten material and intersubjective possibilities for social exchange and identity formation. Collectively, both ideological and semiotic inquiries call upon readers to consider the past, present, and futuristic “chronotopes,” or narrative emplotments of “space-time-personhood” enacted in city streets.³

Stated otherwise, the methods of interdiscursive ethnohistory can be used to compare facets of official versus “grassroots literacy,” the latter concept developed by Jan Blommaert to refer to the nonelite forms of writing produced by people working outside of elite information economies (2008, 7). By zooming in and out of focus on graphic artifacts that have little to no obvious connection to one another, such as grammar books produced in colonial French India and contemporary street signs in South Asian and diasporic neighborhoods, this method aims to reveal dialectical relationships shaping the spatial scope of colonial and postcolonial narratives and temporal scale of different literacy practices. Additionally, the comparison of transliterated signs across the cities of Puducherry, Paris, and Montreal seeks to elucidate the impact of French cultural imperialism on Tamil in spite of the failed French colonial efforts in South India and relatively minor role of the French language in the historical development of Tamil print culture. The political, economic, religious, and kin-based connections that do exist between francophone Tamil communities, even with the lack of political motivation to standardize translation and transliteration practices, is suggestive of a Tamil francophonie. I define Tamil francophonie by its potentiality to become a transnational space where persons who cross boundaries to migrate, undertake pilgrimages, and pursue business ventures all value literacy in French and Tamil.⁴ However, my analysis of the archival, textual, media, and ethnographic sources produced in this space recovers traces of colonial and postcolonial contestations hidden in typographic and orthographic forms and exposes globally interconnected and dynamic processes of sociolinguistic differentiation.

Different anthropological traditions engaging in the semiotic analysis of sociocultural life have recognized that material artifacts, including outdoor signs, communicate not just in a heuristic fashion but also as living signs. Cultural anthropologist E. Valentine Daniel writes that “institutions can generate an agentive moment as long as they contain within them the signs of a human being” (1996, 190), using the argument to explain how houses in Tamil Nadu act as agents to orient Tamil persons to authenticating soils (1984). Similarly, research on animation by linguistic anthropologists recognizes the ritually grounding, pragmatic work involved in transforming the potentiality of signs based on the interpretation of likenesses into those instead seen as being contiguous (Manning and Gershon 2013; Ball 2014). If the semiotic mechanism of ritual animation is to collapse the “separation between actors and events located elsewhere … by bringing them into spatiotemporal contiguity” (Ball 2014, 168), given the right institutional conditions, language and semiotic ideologies, and interactional practices, especially those of a ritual nature, communities with incommensurate claims to genealogy and history in the Tamil-speaking world could recognize themselves in the signage as a veritable Tamil francophonie.

Yet the pragmatic work needed to animate graphic artifacts into signs of embodied agency relies first and foremost on individuals being socialized into multilingual repertoires inclusive of different registers of Tamil, French, and English. In francophone cities where local laws and social norms advise residents against speaking or writing in the global lingua franca of English, one can presume that the street and other outdoor signs that foreground or background English or French function as indexical icons (i.e., sinsigns; see PWP, 102, 115–18) diagramming the contours of a spatiotemporal imaginary (i.e., chronotope). When these signs juxtapose and reconstitute the political space of la Francophonie and cultural realm of Tamilagam out of bifurcated colonial and postcolonial periods, graphic artifacts of translation and transliteration, whether affixed onto street signs or forgotten in obscure colonial handbooks, can potentially speak to distant or proximal addressees across community boundaries. Rather than literature or politics, the defining feature of the resultant Tamil francophonie is the use and occasional mixing of Tamil and French (and English) in both colloquial and literary registers (Das 2016).⁵

Understanding addressivity is thus essential to developing a method of interdiscursive ethnohistory. Although Erving Goffman developed a classification of participant roles as turn-taking conversational structures ([1955] 1972), this classification has been reformulated (Irvine 1996), most recently by Michael Lempert (2011) and Sabina Perrino who reimagine addressees instead as structures of addressivity “to explore how micro-textual forms of temporal semiosis articulate with sociologically inflected spacetime … [to] often remake the very world in which they occur” (Lempert and Perrino 2007, 206; see also Agha 2005a). Although this statement focuses on the mobility of discourse, Lempert and Perrino recognize that “immobile” discourse also “fails” to “speak” across spatiotemporal boundaries and certain constraints impede de- or recontextualization. Highlighting the impasses or currents in diasporic flows and underscoring the reification of sociolinguistic categories in the communicability of signs helps to explicate whether conventional or idiosyncratic transliterations in the form of typographic and orthographic choices on signage act as connectors between colony, metropole, and diaspora across the Tamil francophonie.

The Making of a Transliteration Standard in Pondichéry/Puducherry

Religious and imperial rivalries in colonial India subsumed European debates about the visual representation (i.e., translation and transliteration) of textual artifacts in Tamil. Sixteenth-century Jesuits considered Portuguese, the lingua franca of Catholic missionaries in Asia, as the most perfectible and suitable for translation of all languages due to the close influence of Latin (Županov 1998; Xavier and Županov 2015). Tamil was instead deemed “theologically deficient,” “phonologically barbaric,” “laborious,” “difficult,” and prone to errors caused by all manner of mispronunciations and idiosyncratic writing practices (Xavier and Županov 2015; Das 2017). By the nineteenth century, European writers had adopted a practice established by the prominent Italian Jesuit missionary, C.G. Beschi, for transliterating Indian vowels using only Italian vowel forms and Indian consonants by employing English ones instead. Yet when British Orientalists in colonial Madras (Trautmann 2006) began to regard colloquial Tamil and classical Tamil as two different languages, Low Tamil and High Tamil, they relied increasingly on English phonology for the transliteration of Low Tamil, while turning to the Italian alphabet only for High Tamil.

Missionaries and Indologists in early to midcolonial French India were exceptional in generally eschewing such practices of transliteration. For example, the lieutenant of a French battalion of cypahi (Indian infantry) soldiers, Amédée Blin, and the two abbots running the Missions Etrangères de Paris (MEP) press in Pondichéry, Louis-Savinien Dupuis and Louis-Marie Mousset, wrote bilingual dictionaries by printing the Tamil entries in Tamil script and French entries in Roman script. Although Mousset and Dupuis shared the scholarly opinion of their British counterparts in Madras concerning the recognizable differences between High and Low Tamil (Das 2016, 2017), their decision to combine entries from both registers into a single dictionary obviated the need for creating different source-language orthographies (1895, xxii).

Efforts to codify a French system for the transliteration of Tamil into Roman did not begin in earnest until M. J. Baulez of the Mission Apostolique Vellor printed in 1896, with the permission of the Catholic MEP press in Pondichéry, Méthodes de Tamoul Vulgaire, a handbook for teaching French colonial officials in India how speak colloquial Tamil by learning to write.⁶

Tel est le plan de cette méthode. Parler d’abord, puis étudier. Les versions et les thèmes qui se trouvent après chaque leçon ne sont point un travail: c’est une conversation formée des mots déjà appris. Tous se tient et s’enchaîne. Une leçon prépare à la leçon suivante, et celle-ci ne fait que développer la précédente. (Baulez 1896, 1)

[Such is the plan of this method. First speak, then study. The versions and themes found after each lesson are not at all exercises; they are a conversation formed from the words already learned. Everything stands on its own and is connected together. One lesson prepares for the next lesson, and this lesson develops the previous one.]⁷

By no means an advanced linguist, at one point Baulez confused Gujarati with other Dravidian languages, even though this had been proven otherwise in 1812 by Francis W. Ellis (Trautmann 2006). Nonetheless, after outlining his literacy-based method of language self-instruction in the preface, he proceeded to describe the totality of 30 Tamil letters—12 vowels and 18 consonants—and list them without further explanation, in addition to the corresponding vowels and consonants in the French Roman alphabet. He distinguished the long vowels with the circumflex diacritic (^), rather than doubling the letter, the customary practice favored by British printers, establishing the French as among the few Europeans in India to favor diacritics in printing. He clarified the pronunciation of உ by spelling it as “ou” [u], ஏ as “ei” [eː], and ஔ as “aou” [aʊ̪].

Most of Baulez’s consonants feature a unique combination of lower- and/or uppercase letters to transliterate sounds nonphonemic in French. Thus, the palatal nasal ங [ŋ] is transliterated as “gn,” rather than “ny” as in English. Sanskrit-origin consonants using the Grantha script, such as the voiceless retroflex sibilant fricative ஷ [ʂ] transliterated as “ch” rather than “sh,” and the voiced palatoalveolar sibilant affricate ஜ [d͡ʒ] transliterated as “dj” rather than “j,” emphasize a French-source phonology. Finally, the retroflex approximant ழ [ɻ], which is one of the most recognizable features of literary Tamil although pronounced differently in colloquial Tamil in different regions, is referred to as lingual gras and glossed as “g” in “page,” represented as capital letter “J” or “L.”

A subsequent grammar written by P. Lap, Abrégé de la Grammaire Française-Tamoule, published in Pondichéry in 1904, included a similar standard for the transliteration of Tamil vowels and consonants into Romanized French, presented in the text in tabular form and without metapragmatic commentary (fig. 1). While maintaining Baulez’s practice of transliterating Tamil vowels by using the circumflex diacritic, the Abrégé dropped the use of capital letters and devised new graphemes for consonants such as the alveolar trill ற [r], retroflex nasal ண [ɳ], retroflex stop ட [ʈ], retroflex lateral approximant ள [ɭ], and retroflex approximant ழ [ɻ], which shares the symbol “lh” with the retroflex lateral approximant, a redundancy that perhaps emphasizes the French view of a pragmatic link between literary and colloquial Tamil. Included in the second edition of Dictionnaire Tamoul-Francais are other idiosyncratic usages, such as the occasional use of “bh” for writing the voiced bilabial stop ப [b] in “bharttâ” (substitution) (1895, 349) and “ph” for the voiceless bilabial stop ப [p] in “phânita” (raw cane sugar; 377). There is no phonetic condition to explain this nonsystematic representation of bilabials as the aspirated stops [bʰ] and [pʰ]; these sounds are not phonemic in Tamil in the same way as they are in Telugu and Malayalam, the two Dravidian languages with more pronounced Indo-Aryan linguistic features due to contact with Sanskrit.

Figure 1. Transliteration of vowels and consonants from Tamil to French (Lap 1904, 7)

The most extensive effort toward articulating conventions for transliterating Tamil into French was authored by Julien Vinson in Manuel de la Langue Tamoule (1903). Extensive commentaries accompanied a lengthy section devoted to describing the Tamil alphabet and advising how to transliterate it phonetically into Romanized French. For example, Vinson explained that the choice of “J” for ழ is rooted in a local style of pronunciation associated with the residents of Pondichéry, Karaikal, and Tanjavur, whereas the people of Madras favor [y] as the pronunciation and those in Madurai favor [ɭ]. Vinson also paid close attention to phonetic variation permitted by transliteration conventions, such as the use of ஃ for different types of prosodic vowel modifications. Finally, he provided abundant examples of Tamil-to-French transliteration, showing the visual effects of diacritic use.

Altogether, these late nineteenth- and early twentieth-century attempts to create and disseminate conventionalized forms of phonetic transliteration from Tamil to French set the colony apart from other major printing centers in South India due to the willingness to endorse diacritics and work around the imprecise boundaries between colloquial and literary varieties of Tamil. Taking pride in its typographic innovations and independent printing tradition, the MEP press requested the assistance of the colonial government in Pondichéry in publicizing bilingual books and translations and promoting their circulation outside of the Indian region. In 1865 the governor of French India wrote on behalf of Dupuis and Mousset to the minister of the navy and colonies in Paris to inquire into whether French and Tamil books could be useful to the libraries in the metropole and plantation colonies of Guadeloupe, Martinique, French Guiana, Réunion, and Mauritius, where indentured laborers of Tamil descent had settled from 1830 to 1920. Although there is little trace of such books existing in these remote parts, the colonial record does indicate that the governor of French Guiana ordered a medium shipment of books from Pondichéry in 1863 (Das 2017). Most books from the MEP press remained, however, in the Indian cities of Pondichéry and Karaikal, where mostly high-caste Indians educated in French schools, some claiming mixed-race ancestry as gens de couleurs, were literate in both languages (Carton 2012). Later, after Pondichéry was ceded to India between 1957 and 1963, many of these elites chose French citizenship and migrated to Southeast Asia, France, and, recently, Montreal.

Today, signs with French toponyms, surnames, and cartographic features memorializing Pondichéry’s colonial legacy account for the French-to-Tamil translations/transliterations found in the streets of Ville Blanche, where the occasional Indian surname in Tamil script is also transliterated into Roman script. Surnames refer to major players in Indian Ocean history, including, on the French side, governors-general François Martin, Joseph François Dupleix, Jean Law de Lauriston, Eugène Desbassayns de Richemont, Pierre Benoît Dumas, and Bertrand François Mahé de la Bourdonnais; naval commander Bailli de Suffren; scientist Jacques Surcouf; abbot Louis-Marie Dupuy (also Dupuis); and writer Romain Rolland; and on the Indian side, Dewan Kandappa (also Candappa) Mudaliar, architect of the Vedapureeswarar Temple; Ananda Ranga Pillai (also Rangapoulle), interpreter to Dupleix; and Jawaharlal Nehru, prime minister of India.

Puducherry’s street signs mostly display transliteration conventions similar to earlier ones documented in Baulez’s, Lap’s, and Vinson’s manuals and handbooks. For example, the surname “Gilles” is transliterated using the retroflex approximant ழ [ɻ] for the word-initial consonant “g” [ʒ] (fig. 2). Major departures can be witnessed, however, in the absence of capital letters for writing retroflex consonants and diacritics for long and short vowels. Thus, முதலியார் transliterated as “Moudaliar” represents the short vowel உ as “ou” but does not include the circumflex on the “a” to indicate long vowel ஆ in the last syllable. Some Tamil surnames that are transliterated into Romanized French also feature idiosyncratic orthographies not found elsewhere, such as the exclusively French spelling of Ananda Ranga Pillai’s surname as “Poulle” [pulle] (fig. 3), likely reflecting a regional pronunciation of the Vellālar caste name, which is seen in variants such as “Pillai” in South India, “Pillay” in South Africa, and “Pulle” in Sri Lanka. Some French words are translated directly into Tamil: “rue” as வீதி, “cathedrale [sic]” as மாதா கோவில், and “petit canal” as சின்ன வாய்த்தால். A nonpurist attitude also prevails in the liberal use of Grantha scripts, such as ஹ and ஜ respectively representing the non-native phonemes [h] and [dʒ], whereas the transliteration of [f] in “Surcouf” as ப or [p] can be attributed to the much later use of the grapheme ஃப, and not a preoccupation with purity.⁸

Figure 2. A street sign showing French-to-Tamil transliteration, Puducherry, December 2018

Figure 3. A street sign showing French-to-Tamil transliteration, Puducherry, December 2018

Orthographic choices on street signs, however, suggest a historical shift or inversion in the language ideologies informing the directionality of transliteration from Tamil to French in colonial Pondichéry to French-to-Tamil in postcolonial Puducherry. Throughout the eighteenth and nineteenth centuries, extended contact between French and English in Tamil country was reinforced by successive waves of conquest and transitionary periods of British rule in colonial French India. Ultimately, three codes have emerged to influence writing and printing by the French in Tamil in the contemporary period. Since the French in anglophone India did not aspire to universalist projects of knowledge production through their printing but instead sought to preserve the legacy of their embattled colonial presence, they incorporated multiple colloquial and literary Tamil registers not easily distinguished as either Low or High Tamil into their transliterations. Nonetheless, as the functionality of the French language in Pondichéry faded in the postcolonial period, French script downshifted from alphabetic symbols to become ornamental icons as printed street signs in Puducherry struggled to embody recognizable signs of a Tamil francophonie.

The Nationalist Politics of Transliteration in Montreal

Home to more than 30,000 Tamils of mostly Sri Lankan but also Indian, South African, Mauritian, and Malaysian national origins, Montreal’s diaspora may not comprise the largest Tamil diaspora in the world but it is significant in terms of its strategic location close to Toronto, which is the largest diasporic city of Tamil speakers. In fact, two of Montreal’s most populous and visually distinctive Sri Lankan Tamil neighborhoods, Parc-Extension and Côte-des-Neiges, are known by Sri Lankan Tamil refugees from around the world as “Little Jaffnas” (Daniel and Knudsen 1995, 175–76). These neighborhoods participate in a dynamic of fission to recreate competing residential, business, and religious communities along the Toronto-Ottawa-Montreal corridor (Das 2016). Crucial to the expression of this rivalry are signs affixed on the storefronts of restaurants, boutiques, temples, and other buildings that cater to customers and patrons of Sri Lankan, Indian, Caribbean, African, and other ethnicities, while also addressing government officials and nationalist leaders overseeing choices of code, script, and orthography on signs. Storefront signs use multiple participant frameworks laminated through the use of standard and nonstandard writing conventions to send overt and covert messages about how businesses are legitimately making a profit, aspiring to fame, and demonstrating respectability. Acting as agents, such signs also bring to life the treasured qualities of hometowns for immigrants and refugees.

Among the different South Asian diasporic communities established in Montreal, Sri Lankan Tamils have capitalized the most in exploiting the visual affordances of their written language to selectively highlight or conceal competing postcolonial allegiances on the facades of storefront signs (Das 2016). In general, Sri Lankans who identify as heritage speakers of Tamil primarily value the use of French in its capacity as Québec’s official language and English as Canada’s other official language. They maintain ties with other diasporic communities through their migration, business, and marriage networks, including in Paris where French and Tamil are also spoken. These language-based connections are a matter of legal necessity too. Since 1976 the Charter of the French Language has mandated that children of immigrants attend public schools in French and business owners print storefront signs with bigger, bolder, and more front-and-center French graphemes. Compared with Ville Blanche in Puducherry, where the use of French is increasingly ornamental and Tamil is printed in Tamil script only, on Montreal’s storefront signs, Tamil graphemes, which are printed in both Roman and Tamil scripts, are subjected to neighborhood scrutiny and disciplinary action for failing to meet the province’s legal requirements. These nationalist policies translate into outdoor signs communicating to differently scaled addressees, proximal and distant (LaDousa 2002; Frekko 2009), offering insider information about provenance and loyalty to one’s identity and politics as being Hindu, Catholic, Protestant, or Muslim, high or low caste, LTTE or unaffiliated.

By comparing changes in storefront signs in Parc-Extension and Côte-des-Neiges over the span of twelve years, starting from 2006 a few years before the end of Sri Lanka’s civil war until 2018, I identify subtle changes indexing the maturing presence and outward-looking stance of Sri Lankan Tamil businesses in Montreal. Restrictions put in place on the political activity of the LTTE and World Tamil Association in Canada in 2006 and again after the end of civil war in Sri Lanka in 2009 ushered in new political conditions, leading to fewer refugee and greater family reunification claims (Jedwab 2005).⁹ Côte-des-Neiges, the entry point for many refugees living in multifamily apartments, has become less residential than a decade earlier, after many families have moved to suburbs or less crowded neighborhoods in Montreal. The business district in Côte-des-Neiges, however, has remained intact, with only five fewer establishments in 2018 than in 2006 (see tables 1 and 2).

Table 1. Tamil Shops along avenue Victoria in Côte-des-Neiges in 2006

Table 2. Tamil Shops along avenue Victoria in Côte-des-Neiges in 2018

The turnover of these establishments demonstrates that, whereas grocery stores continue to be popular in Côte-des-Neiges, more upscale restaurants, boutiques, and beauty salons have begun catering to a wealthier clientele now frequenting Tamil-owned businesses. Moreover, compared with the modest display of storefront signs printed in Tamil in 2006, in 2018 the script itself is often more prominent (fig. 4) and displayed with greater flair in typeface. This quality of “doing style” (Nakassis 2016) differs from the plain Roman fonts used on the nearby Filipino, Vietnamese, and Indian storefront signs, including on a self-identifying Indian Tamil restaurant called Thanjai located on the intersecting street of Van Horne, publicizing food in Romanized Tamil. In Montreal, Indians have embraced “Spoken Tamil” as their heritage language and do not invest in the display of expertise in literacy, which falls under the rubric of “Written Tamil” and jurisdiction of Sri Lankans (Das 2016). Hence, no Indian Tamil businesses actually print their signs in Tamil script. Further driving the ethnolinguistic differentiation of Indian and Sri Lankan Tamils in Montreal are ideologies stigmatizing Sri Lankans as gang members, terrorists, and culturally unassimilable persons. Since many Quebec residents assume that Indian Tamils and Sri Lankan Tamils constitute a single ethnolinguistic group on account of both groups using the same heritage language, Indians challenge this language ideology by overemphasizing the grammatical differences between Indian and Sri Lankan varieties, glossed as Spoken and Written Tamil, whereas Sri Lankans overemphasize the literary and ancient qualities of their language (Das 2008, 2016). The fact that these verbal differences are reproduced graphically through the presence and absence of Tamil script on the signs of Sri Lankan and Indian Tamil businesses, respectively, illustrates pervasive colonial and postcolonial tensions shaping diasporic neighborhood formation.

Figure 4. A storefront sign showing French-to-Tamil transliteration, Côte-des-Neiges, December 2018.

The neighborhood of Parc-Extension differs from Côte-des-Neiges based on the proximity of a major Hindu temple, Sri Durkai Amman, and a Catholic church, Our Lady of Deliverance, in addition to a thriving business district, residential area with apartments and duplexes, a school offering Tamil heritage language classes and a large park with cricket fields. On a subway map I notice the words “Little India” scribbled on top of the Parc-Ex stop, a misappellation since most Parc-Ex residents of South Asian descent are actually Bangladeshis, Pakistanis, and Sri Lankans. Compared with the storefront signs featured on Côte-des-Neiges businesses, in Parc-Ex the religious and political ideologies influencing the choice of orthography and script there are more discernible to viewers literate in Tamil and French. Adhering to the values of linguistic purity and religious nativism enshrined in the philosophy of Saiva Siddhantism, for example, the Sri Durkai Amman temple displays an outdoor sign printed in Literary Tamil in Tamil script that observes sandhi rules and provides the French and English translations in Roman script. Comparing this temple sign with a commercial sign featuring the same goddess’s name underscores differences in ideological stance. Rather than faithfully transliterating the voiceless velar stop [k] as “k,” the commercial sign instead transliterates it as voiced velar stop [g] to print the name “Thurga” using a “g.” Also, the commercial sign employs the அ grapheme, transliterated as “a,” to print the more familiar English name, Thurga, compared with ஐ used by the temple to represent the final vowel as “ai” [aɪ], a purist Tamil spelling. Only Marché Thurga is printed using the Grantha script ஷ to transliterate the “ch” in “marché” as the non-Dravidian phoneme [ʂ] (Das 2016, 162–63). Recall that French colonial missionaries in the nineteenth century insisted that ஷ be translated as “ch” following the French alphabet convention, rejecting the more common use of “sh” favored in English texts.

Storefront signs in Montreal’s Tamil diaspora sometimes rely on English to display their familiarity with standard and nonstandard varieties of Quebecois French, yet this controversial presence of English is obscured by the Tamil transliterations. Two spellings of the same name, “Marché Jeevini” on the online business directory and “Marché Jeeveni” on the storefront sign (fig. 4), demonstrate the muted presence of English. The first sign corresponds to a French-based orthography in which the short vowel [i], represented as இ in Tamil, is transliterated as “i.” This sound is somewhat similar to the [iː] of Standard French yet better resembles the [ʏˑ] of nonstandard Quebecois French. The second spelling, based in English, transliterates the long vowel [iː], represented by the grapheme ஈ, instead as “ee,” and the short [i], represented as இ on the sign, as “e.” When read aloud in standard or nonstandard Quebecois French, the Roman grapheme elicits [ə], a sound very different from both standard colloquial and literary Tamil pronunciations of “e.” The grocery store sign also omits the voiced uvular fricative [ʁ] in “marché” by transliterating it as மாசே [māce], producing, when read in Tamil, an irregular utterance approximating a highly stigmatized pronunciation when spoken in nonstandard Quebecois French.

No business can function long in Montreal without adhering to the provincial language policies that mandate accurate French spellings and standard grammatical forms. In 2006, when I took a photograph of a daycare in Parc-Ex for Tamil children called “Angel Daycare,” I noticed the Tamil name was transliterated phonetically from English as ஏஞ்சல் டேக்கெயர் and translated into French as “Angel La Garderie Du Bon Berger.” However, it would be misleading to label this translation as French, due to lexical and syntactic anomalies, since it should instead read “La Garderie de l’Ange du Bon Berger.” In a Google Image observed in June 2019, I saw that the daycare had been renamed Centre Éducatif du Parc, and translations and transliterations into Tamil were conspicuously missing from the storefront sign, even though decals of the former name, “Angel La Garderie du Bon Berger,” were still pasted on the front door. In negotiating the many linguistic rivalries between English and French, standard and nonstandard Quebecois French, and literary and colloquial varieties of Tamil in Montreal, the printer of the new sign perhaps sought to tailor its message to multilingual viewers navigating adversarial sociolinguistic worlds.

Regardless of their legal status, most of these storefront signs reveal embedded structures of addressivity that underscore the rapid demographic growth and rising fame of Sri Lankan Tamil neighborhoods in Montreal. There, unlike in Puducherry, residents can boast of expansive opportunities for housing, employment, and entertainment, as well as heritage language schools for future generations of francophone Tamils also compliant with provincial laws protecting French. Elsewhere, I have described these signs as anchoring neighborhoods in familiar scenes of home and enabling them to compete for global recognition during the transient period of civil war in Sri Lanka (Das 2016). Here, I instead analyze these storefront signs for their potential to speak across communities of francophone Tamils. Drawing on Silverstein’s concept of indexical order, I argue that, at the n ^th indexical order, the presence of Tamil script is interpretable as a graphic artifact. At the n ^th+1 indexical order it instead becomes iconic of the presumed purism and ancient origins of Literary Tamil, due to the influence of Dravidian and Sri Lankan nationalist ideologies, and at the n ^th+2 indexical order, due to ethnolinguistic classifications popular in Montreal, emblematic of a local Sri Lankan community, seen as the speakers and custodians of a purist “Written Tamil” (Das 2016). I also propose an n ^th+3 indexical order in which outdoor signs entail contingent connections between Montreal and Paris, but less successfully so with the streets of Puducherry.

Divergent strategies for transliterating French into Tamil in Montreal suggest a lack of familiarity with (post)colonial printing traditions in Pondichéry/Puducherry. One discontinuity is the absence of diacritics to differentiate between long and short vowels printed in Roman script. Another case is the word-initial consonant [d͡ʒ] in “Jude” being transliterated as [y] producing யூட், thereby following a Sri Lankan or Madrasi convention rather than using the Grantha script ஜ as prescribed in French colonial manuals. On the other hand, there is continuity in the use of a French-based orthography to represent nonphonemic sounds in Tamil, such as the [ʂ] in “marché” as ஷ (fig. 4), written like this on all but one grocery store in Montreal, which instead uses ச [s] to adhere to the ideology of linguistic purism (fig. 5). Such divergent choices belie the practical difficulties in enforcing transliteration standards given the historical and intertextual gaps shaping the historical circulation of texts among francophone Tamils.

Figure 5. A storefront sign showing French-to-Tamil transliteration, Parc-Ex, April 2006

In contrast, the residents of Montreal and Paris have cultivated tangible and enduring connections fostered through migration, tourism, business, kin relations, and religious rituals. Narrating with fondness and nostalgia her prior life in Paris is Marianne, a Pondichérian Tamil who moved to Canada as a teenager and often visits Catholic establishments operated by French nuns in Montreal to relive her feeling of spiritual connection to Tamil Catholicism rooted in France. Pavalan, who enjoys visiting his cousins in Paris, instead describes the French culture there as “too strong” for him to fully express his Hindu identity and prefers living in Montreal, where he can speak French, Tamil, and English and aspire to found a sports apparel business while enjoying his role as the youngest board member of the Thiru Murugan Temple. Mala, who also has cousins in Paris, goes there to cultivate expertise in International French used with her friends of Arab descent emigrating from Paris to Montreal (Das 2016). Conversely, opportunities for Parisian Tamils to imbibe in the vibrant scenes of Montreal’s francophone Tamil community revolve around the religious festivals inaugurating or blessing churches and temples, along with the marriages and business transactions that accompany these pilgrimages. All three types of events produce mass quantities of grassroots literacy in the form of flyers, banners, cards, and other print media. When viewed by visitors and migrants, these textual artifacts animate signs of contiguity and downplay the geographic and cultural distance between Montreal and Paris to attract new residents and tourists to Canada.

The Multiple Lives of Transliterations in Paris

The Tamil community in France, immersed in a whirlwind of competing legacies of colonial and postcolonial migration, labor, and nationalism, includes approximately 50,000 Tamils of Indian Pondichérian descent (Dassaradanayadou 2007), somewhere between 50,000 to 70,000 Tamils of Sri Lankan descent (Dequirez 2007; Goreau-Ponceaud 2009), and other Tamil groups originating from the Antilles and Mascarene islands. From May through July 2015 as a visiting scholar at the École des Hautes Etudes en Science Sociales, I gave a series of lectures about my research on francophone Tamils in Montreal. Audience members of Sri Lankan and Pondichérian Indian descent, raised in Paris to be bilingual in French and Tamil, were fascinated to learn of the experiences of francophone Tamils in Montreal. I also visited the neighborhood of La Chapelle, originally called “Little India” and more recently renamed “Little Jaffna,” at the heart of this urban francophone diaspora. Sandwiched between the stations Gare du Nord and Gare de l’Est in the tenth and eighteenth arrondissements, La Chapelle has been the primary entry point for Tamil immigrants since 1956, when the end of French rule in India afforded the opportunity for Pondichérian Indians with French citizenship to pursue free higher education studies in Paris (Dassaradanayadou 2007). Later, Sri Lankan Tamil refugees, mostly from the northern province of Jaffna but also Colombo, began to arrive in Paris in the early 1980s. This migration pattern continued through the 1990s and early 2000s. Both of these communities have founded small establishments, such as grocery stores, beauty salons, clothing and fabric stores, video stores, and cultural associations, to recreate the feeling of home and facilitate integration into Parisian society.

In a study of this neighborhood, a young woman of Pondichérian Indian ancestry is cited as saying that La Chapelle recreates familiar scenes of her hometown of Puducherry, relaying to her interviewer that “Gare du Nord, c’est mon quartier, à une époque j’y allais tout le temps, tous les deux jours. C’était vraiment Pondi, c’était la rue Nehru pour moi, à chaque fois que je sortais j’étais bien. C’est un lieu de rencontres, je rencontre beaucoup de gens” (North station, that’s my neighborhood, at one time I used to go there all the time, every two days. It was really like Pondi, it was Nehru street for me, every time that I went out, I felt good. It’s a meeting place, I meet a lot of people) (Dassaradanayadou 2007, 78). After the first boutique opened in 1982, Sri Lankan Tamil establishments began to visually take over the neighborhood, as an influx of Sri Lankan refugees arriving in the 1990s added to the grocery stores, restaurants, and fabric stores previously established by Pondichérians. Currently, services such as travel agencies, photographers, hardware stores, and real estate agents cater to Sri Lankan Tamil immigrants, in addition to two Hindu temples and Tamil schools teaching language, dance, and religion (Dequirez 2007). Complicating this shift are the economic difficulties in maintaining a historically traditional Tamil clientele, leading some businesses to expand outside of their base by offering “des pommes et du vin” to attract other Parisians and tourists (85).

Different migrant cohorts have impacted the visual composition of storefront signs in La Chapelle either by foregrounding English and backgrounding Tamil and French or by making explicit references to ethnonational origins. A rudimentary typology of storefront signs identifies signs written entirely in French that emblematize the exoticness of Indian food and culture (e.g., Le marché exotique, Allô Tandouri!) to entice non–South Asian tourists and customers in Paris (Goreau-Ponceaud 2013, 12). Another type of sign also written only in French evokes regional names, such as Océan Indien and Jaffna Boucherie, to conjure images of faraway places from which clients of different national origins have migrated. A third type of sign, written in English (fig. 6) but also sometimes in French, includes toponyms, surnames, and the names of deities on storefront signs that address customers of different religious and caste backgrounds (Goreau-Ponceaud 2013). Few signs are printed in Tamil. One exception is Balavinayagar Cash & Carry, transliterated from பாலவிநாயனர், the name for a popular temple dedicated to Ganesh in Chennai.

Figure 6. Storefront signs that are in English only, Paris, June 2015

Diasporic businesses with a lack of consistent preoccupation with linguistic purism and changes in ethnonational composition are the two features shaping this linguistic community of francophone Tamils, united by their adoption of the prestigious International French standard yet also open to contact with written English. Three Tamil establishments located next to one another—the Muslim-owned halal butcher shop Boucherie Gabarina; the Indian restaurant Restaurant Indien Muniyandy Vilas; and the joint Indian and Sri Lankan restaurant Restaurant Annachi—display these stances through a mish-mash of translations and transliterations printed in French, English, and Tamil (fig. 7). Boucherie Gabarina’s sign mimics the style used in Puducherry of first transliterating the proper name, in this case “Gabarina,” and then translating the common noun, in this case boucherie. Montreal business owners do not adopt this practice. Instead, grocery store signs there transliterate marché directly into Tamil script, rather than translating it as “maḷikai viyāpāram” மளிகை வியாபாரம், which means grocery store in Tamil. Although the word for “restaurant,” or “uṇavakam” உணவகம், is used for both Restaurant Annachi in La Chapelle and Restaurant Jolee in Côte-des-Neiges, a different word is used for Restaurant Janani in Montreal, விருந்தகம், meaning “guest house.” On Parisian signs, some store names transliterated from English to Tamil omit French entirely. Thus, the storefront sign for the barber shop New Prahba Saloon only provides the transliteration (நியூ பிரபா சலூன்) in Tamil script. Also, the Romanized version of “Prabha” is written in a nonpurist orthography, yet when transliterated into Tamil it switches into a purist orthography, using ப [p] to represent the non-native phoneme [bʰ] in Tamil and observing sandhi rules for printing “pra” as பிர [pira].

Figure 7. Storefront signs that transliterate proper names and translate common nouns, Paris, June 2015.

Celebrating the diversity of ethnonational, religious, and caste identities and business approaches, these storefront signs in La Chapelle rely less on the visual aesthetics of Tamil script and more on the prestige of International French and normativity of spoken Tamil to address clients and entice new visitors. In comparison, the lack of prestige attached to Quebecois French in Montreal compels viewers there to look toward France for the grammatical standard. Without greater value being placed on Tamil and French literacy in La Chapelle, this neighborhood will not be able to compete in the future with printing practices in Montreal, where residents produce and circulate texts in Tamil script throughout the world. The relative paucity and multiple lives of Tamil script on Parisian signs undermine its communicability abroad and likely constitute a transitory semiotic ontology between the ornamentality of French in Puducherry and iconicity of Tamil ethnonational identity in Montreal.

Conclusion

The interdiscursive ethnohistory of the street, storefront, and other graphic artifacts found in four francophone neighborhoods is based on linguistic, ethnographic, and archival analyses of the disparate sociohistorical contexts in which businesses and organizations either promote or devalorize literacy and printing in Tamil, French, and English. Whether these signs project the image of an interconnected realm (which I have labeled “Tamil francophonie”) depends on how embedded structures of addressivity animating graphic artifacts lead to encounters between residents originating from Montreal, Paris, and Puducherry. I have argued in this article that the translations and transliterations in Tamil, French, and English driving the recalibration of chronotopes bring three francophone cities with separate legacies of Tamil settlement into potential contact with one another. Embodying the present, Paris provides the grounds for reproducing the linguistic community through residents’ adherence to the prestige of the International French standard. Constrained by the stigma of Quebecois French, Montreal’s neighborhoods instead embody the future of the diaspora, emphasizing the vibrancy of its entrepreneurial and printing activities, including the presence of a heritage language industry invested in grassroots literacy. Although street signs in Puducherry feature mostly ornamental displays of French and fail to convince francophone Tamils of a shared history, they nonetheless offer opportunities to connect with an imagined past in which Tamil and French coexist in colonial handbooks and postcolonial street signs.

One can also discuss how intersubjective and material connections between these three cities are manifested in semiotic terms. Attempts to produce a standardized system of Tamil-to-French transliterations in colonial Puducherry failed to anticipate the need for French-to-Tamil transliteration conventions in the postcolonial context. Additionally, as French declined in use in South India throughout the nineteenth and twentieth centuries, an “inversion” of indexical orders (Inoue 2004) there likely occurred in conjunction with the long-term demographic and political changes that have since redrawn the map of Ville Blanche from an elite neighborhood for French citizens into more of a tourist attraction. Similarly in La Chappelle, the shift from Pondichérian to predominantly Sri Lankan-owned businesses corresponds with a shift in migration patterns. However, due to a lack of preoccupation with heritage languages in Paris, Romanized French and English continue to be used on most print media, making transliterations into Tamil less necessary or noteworthy. It would be revealing if pictures of storefront signs taken before Sri Lankan businesses took over in the 1980s showed an inversion in the values of scripts and codes, with greater emphasis placed on English and sandhi rules expected after 1983 and greater use of French expected prior to 1983. Taking into consideration the relative stability of the ideological regimes impacting the two national communities comprising Montreal’s Tamil diaspora, one does not see inversion. Instead, the indexical order in Montreal preserves the interpretation of an iconic link between written and heritage languages and ethnic identity, while also allowing for contingencies to emerge in the animation of diasporic life. The more stylized and outward looking of Sri Lankan businesses experiment with typography to aspire for fame and profit yet still uphold their claim to preserving Written Tamil by producing transliterations in Tamil script.

This generative interplay between the embedded structures of addressivity indexed by outdoor signage in Puducherry, Paris, and Montreal represents a potentiality, or firstness, rather than an actuality, or secondness, of inter-connected communities of francophone Tamils (PWP, 104–5). Such observations raise questions about the linguistic mediation of social life in globalizing contexts. Can Puducherry’s bilingual heritage in Tamil and French be brought back to life to reconnect with the “living” diasporas located in Montreal and Paris? Can Parisians always rely on the prestige of International French to justify their preeminent role in orienting other francophone Tamils toward this linguistic standard? Can Montreal’s diasporas persist in ritually grounding and performatively enacting the enduring ties that bind francophone Tamils together through their marital, business, and educational activities, despite having to reconcile contentious postcolonial agendas? Does one need a standardized system of Tamil-to-French transliteration to achieve these goals, especially if conventions for French-to-Tamil transliteration remained fluid throughout the colonial period? Since the animation of graphic artifacts takes place both through copresent and absent signs (Derrida 1976), one requires a methodology that can identify which colonial and postcolonial tensions are evident or concealed in the typographic and orthographic forms found across different types of print media. The method of interdiscursive ethnohistory conceptualizes the generativity of semiosis as partaking in a dialectic of dialectics operating at multiple spatiotemporal scales. Identifying the entailments of transliterations between Tamil, English, and French hence locates the reproduction of a Tamil francophonie in the context of diasporic flows, (post)colonial formations, and revitalization projects viewed from city streets.