Trilingual Blunders

Public Signage and Semiotic Ideologies

Scholarship on nation-building in multilingual contexts demonstrates how public signs both reflect and produce social relations between majority and minority groups (Orman 2009). Case studies in conflict-ridden nations show how the linguistic composition of a sign, or the presence or absence of a language from a sign, are implicated in nationalist projects and discourses of identity (Ben-Raphael et al. 2006; Azaryahu 2012). Errors and outright omissions of Tamil content from multilingual public signs commonly occur in Sri Lanka and also in India, Singapore, Malaysia, and Canada.⁵ But these issues are particularly sensitive in the fragile Sri Lankan postwar context.

The semiotic system developed by Charles Sanders Peirce (1839–1914) is useful for exploring linguistic and visual systems in relation to the physical world. Peirce theorizes how representation between the sign and object is mediated by an intepretant, which represents some “further action or mental representation” (Peirce, cited in Parmentier 1994, 5). His widely cited second trichotomy posits relationships between the sign and object as grounded in formal resemblance (icons—a picture of a fire), spatiotemporal contiguity (indexes—smoke signaling a fire), or conventionality (symbols—the word “fire”) (Irvine 1989, 263). Indexes differ from icons and symbols in that “psychologically, the action of indices depends upon the association by contiguity and not upon association by resemblance or upon intellectual operations” (PWP, 108). Linguistic anthropologists have expanded studies of language-planning policies as discourse to explore how people make sense of the indexical relationships between public signs and advertisements and the social and material world (LaDousa 2011; Shankar 2015). An individual’s recognition of iconic and indexical relationships invariably involves some degree of conventionality, or Peircean thirdness, which can be understood as a rule or law that makes different instantiations instances of the same thing (Keane 2003).⁶

Reminding us of Peirce’s statement that icons and indexes in and of themselves “assert nothing” (PWP, 111), Webb Keane discusses how the process of making sense of iconic and indexical relationships involves preexisting notions about how signs function in the world (2003). Expanding upon the concept of language ideologies (Silverstein 1979; Irvine 1989), he defines semiotic ideologies as “people’s underlying assumptions about what signs are, what functions signs do and do not serve, and what consequences they might or might not produce” (Keane 2018, 65). Semiotic ideologies can determine how a person imputes intention to particular acts of signification and how accountability or responsibility is attributed for these acts to particular agents (also see Hill and Irvine 1992; Enfield and Kockelman 2017). They also concern how people posit a relationship between a particular sign and its object (e.g., as arbitrary or necessary) (Keane 2003). Semiotic ideologies show us how beliefs and ideas about signs contribute to “the ways people use and interpret them, and on that basis, form judgements of ethical and political value” (Keane 2018, 67). Key to this concept, semiotic ideologies are not total and complete systems but are processes that are subject to continued reinterpretation and destabilization (Keane 2003).

By analyzing how Sri Lankan Tamil speakers employ disparate semiotic ideologies to interpret trilingual signage errors, I build on ethnographic approaches to media centered on reception and audience (Mankekar 1999; Ginsburg 2002; Larkin 2008). Social media, by virtue of bringing together diverse groups of people on a previously unfathomable scale (Hillewaert 2015), have been crucial in the formation of national and transnational “imagined communities” or publics (Anderson 1991; Spitulnik 2002; boyd 2014). While it does not make sense to sharply differentiate digital from nondigital media (see Tenen 2017), it is important to understand how “technology helps to shape the participant structure brought into being through its use, simultaneously enabling and limiting how communication can take place through that medium, how the communication circulates, and who can participate” (Gershon 2010, 285). This article considers how social media technologies have enabled the creation of forums for individuals to discuss and debate visual semiotic forms as social action. It contributes to semiotic studies of social life by showing how media images as forms of materiality are enmeshed in contemporary transnational political practices.

I begin by looking at the circulation of images of Tamil signage errors on Facebook. In the second part of the article, I analyze the semiotic ideologies employed by the three Sri Lankan Tamil speakers in relation to the views of others expressed in Facebook comments and personal conversations. Finally, I locate my participants’ views with respect to different ethnic, religious, and regional groups in Sri Lanka and discuss their political ramifications. While a small number of Tamils (almost all of whom hail from the north and east) argued that Tamil blunders are evidence of the Sri Lankan government’s malicious intent toward Tamils or Tamil speakers (see above), all of my main participants viewed them as unintentional. However, while two of my participants emphasized the negative effects of the blunders as demeaning and derogatory to Tamil speakers, a third participant invoked a different semiotic ideology by questioning the ability of the posted signboard images to accurately represent the world. She argued that attention to decontextualized images of signboard blunders on social media draws attention away from on-the-ground efforts to achieve language rights in Sri Lanka, efforts that involve services as well as signage. My participants’ and other Tamil speakers’ contrasting interpretations of Tamil signage errors call into question what kind of future is possible for Sri Lankan Tamils following the end of the war, as well as which groups—those living in Sri Lanka or abroad—can and should be involved in creating that future.

My status as a white American female academic who had previously conducted research in Sri Lanka affected how my informants interacted with me. I met my participants during ethnographic fieldwork in Kandy and Colombo (2007–8, 2011), in Sri Lankan and Tamil studies academic networks in the United States and Canada, and on Facebook. I became acquainted with my first primary participant, Arshad (Muslim), through a Kandy NGO in 2007. I met Ruban (Tamil) and Darshini (Tamil) through Facebook discussions on Tamil blunders in 2015, but we were already indirectly connected through mutual friends. While the discussions I present here primarily occurred in English, my participants were aware that I had a high level of proficiency in Tamil as a result of my language study in Sri Lanka, India, and the United States. I am embedded in this account since I took part in some of the Facebook activity, and the questions I asked my informants influenced their responses. Before turning to the media and ethnographic analysis, I discuss current trilingual policies in Sri Lanka in relation to the postwar political situation, as well as the centrality of the internet in the formation of transnational Tamil publics.

Sri Lankan Ethnic Identities and Trilingual Policies

Sinhalas (Buddhist or Christian) make up 74.9 percent of Sri Lanka’s population (see Table 1). They speak Sinhala, an Indo-Aryan language related to the languages of North India. There are several Tamil-speaking ethnic minority groups. North and East Tamils (11.2 percent) have lived on the island for centuries, primarily in the north and east, but also in large cities including Colombo and Kandy (in central Sri Lanka).⁷ Up-Country Tamils (4.2 percent) are descendants of migrants who arrived from South India during the British colonial period (1815–1948) to work as plantation laborers in the central highlands (Daniel 1996).⁸ Members of both Tamil groups are predominately Hindu, with a significant Christian minority. Muslims (or Moors) make up 9.2 percent of the population. They can be traced back to the pre-Islamic seafaring trade between South and Southeast Asia and the Middle East, as well as Arab Muslim mercantile trade in the first part of the seventh century. Muslims live in the north and east and in scattered pockets throughout the south. The majority of Sri Lankan Muslims speak Tamil as a first language and attend Tamil-medium schools, but they ethnically distinguish themselves from Tamils on religious grounds (Thiranagama 2011). This sharply contrasts with Muslims in Tamil Nadu, India, who accept both linguistic (Tamil) and religious (Muslim) identities (Ramaswamy 1997; McGilvray 2008).

Sri Lankan scholars point to the impact of colonial rule in fixing stable ethnic labels to what were much more fluid religious, caste, regional, and linguistic categories (Rogers 1994; Thiranagama 2011). In the mid–twentieth century language-based ethnicity emerged as a primary mode of sociocultural and political identification for Sinhalas and Tamils. During the nineteenth and early twentieth centuries, Sinhalas defined themselves as Kandyan, who came under British rule relatively late, or Low-country, who lived in the southwest coastal region and had more sustained contact with colonial rulers. These regional differences were de-emphasized in favor of a common language-based Sinhala ethnic identity in the 1930s, though sociocultural, linguistic, caste, and class differences among these groups remain relevant. Differences among the various Tamil-speaking groups have been much more openly articulated. Tamils recognize significant regional and ethnic differences between people in the north, east, and Up-Country regions in relation to caste, class, religion, and language (Daniel 1996). In the late nineteenth century, southern Muslim leaders declared Muslims a distinct racial group from Tamils to obtain separate political representation in the colonial government. In the twentieth century, southern urban-based Muslim leaders gradually constructed a pan-Islamic identity, which allowed them to distance themselves from the Sinhala-Tamil conflict. While a Muslim-Tamil political alliance emerged in the northeast for brief periods in the mid- to late twentieth century (McGilvray and Raheem 2007), events that occurred in the 1990s—the violence toward Muslims in the east and the LTTE’s forcible eviction of Muslims from the north—created a rift between these communities (Thiranagama 2011).

Postindependence language and education policies were part of the complex and multifaceted causes of the Sri Lankan ethnic conflict and civil war (Tambiah 1986; Spencer 1990; Thiranagama 2011). Following independence from British rule in 1948, the Sinhala Buddhist-majority government instituted discriminatory policies against the minority groups (Tamils and Muslims), who they believed had received preferential treatment under British colonial rule. One of the most significant of these policies was the Sinhala-Only Act of 1956, which declared Sinhala to be the sole official language of the nation. This act made Sinhala fluency a requirement for all government jobs. While it negatively impacted Muslims, this act was particularly detrimental for English-educated Jaffna (north) Tamils, who, as a result of scarce natural resources in the Jaffna Peninsula, relied on government and professional employment in the south (Tambiah 1986). In addition, a new university policy hurt Jaffna Tamils’ status and future prospects, although it benefited other Tamil-speaking groups (Sørensen 2008).⁹ While postindependence educational policies were successful in improving educational access for all Sri Lankan youth, the separation of students on the basis of language of instruction (Sinhala or Tamil) heightened feelings of interethnic difference and mistrust (Davis 2020).

In the decades after the passing of the Sinhala-Only Act, the government made several attempts to incorporate Tamil into the nation’s language policies. In 1987 the thirteenth amendment to the constitution established Tamil as a co-official language alongside Sinhala and English. But Tamil’s new status remained mostly on paper, and despite the fact that English had played an unofficial role as a link language among elites in the British period, it was not heavily promoted because it was seen as foreign (Devotta 2004). In 2005 the Official Language Commission (OLC), under the leadership of President Chandrika Kumaratunga’s People’s Alliance government (1994–2005), issued a referendum calling for the full implementation of the Official Languages Policy. It pointed out the significant dearth of government officials in the south who could communicate with the Tamil-speaking populace. It also noted that many public signboards were in Sinhala only (Official Language Commission 2006). The OLC and other government bodies created Sinhala-as-a-second-language and Tamil-as-a-second-language training programs for government administrators and police officers throughout the island (Davis 2020).

Sri Lanka’s political climate shifted significantly with the election of Mahinda Rajapaksa as president on November 17, 2005. He quickly sought to eradicate the LTTE by military means. From January to May 2009 the Sri Lankan army underwent an aggressive military campaign to take control of the last LTTE strongholds in the northeast of the island. The United Nations conservatively estimated that 7,000 civilians were killed in this final period of fighting (Thiranagama 2011). The years following the end of the war saw the continued militarization of public life and the proliferation of Sinhala nationalist rhetoric (Goodhand 2012). Rajapaksa emphasized the importance of trilingualism to the postwar reconciliation process, but it is possible that his promotion of a multilingual and united Sri Lanka may have been a strategy to divert attention from his refusal to address the concerns of the nation’s minority groups (Davis 2020). On January 9, 2015, Maithripala Sirisena defeated Rajapaksa. Sirisena, who had served as minister in the Kumaratunga government, cosponsored a United Nations Human Rights Commission resolution to promote “democracy, accountability and human rights in Sri Lanka” and to implement a comprehensive transitional justice process (United Nations General Assembly 2015, 1). He also pushed for full implementation of the Official Languages Policy (Government of Sri Lanka 2018). Many Sri Lankans hoped that Sirisena would lead the nation toward reconciliation as well as acknowledge the war crimes committed by both sides in the final months of the war (Hammer 2016). But the Office of the United Nations High Commission stated in a recent report that the slow pace of transitional justice and the lack of a strategy to address accountability for war crimes remain impediments to reconciliation efforts (UN News Center 2017). The nation faced a new political crisis on October 26, 2018, when Sirisena appointed Rajapaksa to replace the incumbent prime minister, Ranil Wickremesinghe, a move that cabinet ministers deemed unconstitutional. Although the crisis was partially resolved when Rajapaksa agreed to step down on December 14, the prospect of his returning to power left Tamil-speaking minorities fearing for the future (Ananda and Ganeshananthan 2018).

The Sri Lankan Diaspora and the Internet

Sri Lankan Tamil diaspora communities are made up primarily of North and East Tamils who fled Sri Lanka after the outbreak of war in 1983. The largest concentration of Sri Lankan Tamils abroad is in Canada, with the largest population—between 200,000 and 300,000—in Toronto, which was considered the diasporic headquarters of the LTTE during the war (Das 2016). Diasporas are usually defined as groups of people living outside their homeland, but scholars have called for a more analytically rigorous definition of the term (Eisenlohr 2006; Thiranagama 2014). Rogers Brubaker, for instance, argues that diasporas should be treated not as a bounded group but a “category of practice, project, claim, and stance” (2005, 13). In a similar spirit, R. Cheran suggests that Sri Lankan Tamil diaspora communities “cannot be traced to a simple origin or place, rather what creates the diaspora here is a particular kind of imaginary that derives sustenance from nostalgia, pain and loss” (2007, 157).

The relationship between Sri Lankan Tamil diasporas, the LTTE, and “homeland” Tamil populations is highly complex. Tamil diaspora populations provided significant financial support for the LTTE during the war (Cheran 2003; Das 2016). Human Rights Watch (2006) estimates that by the mid-1990s, 80–90 percent of the LTTE’s budget came from overseas sources, with the Canadian diaspora contributing between one and 12 million dollars per year. Some member of the diaspora willingly supported the war while others were subjected to “intimidation, extortion, and physical violence” as the LTTE sought to “suppress criticism of its human rights abuses and to ensure a steady flow of income” (2006, 1). Sharika Thiranagama argues that the Toronto Tamil community’s support for the LTTE was not inevitable, but rather historically contingent, crediting the LTTE with the ability to shape just how Tamil identity was shared away from Sri Lanka (2014).

New virtual spaces were crucial for enabling Tamils in Toronto, London, and other cities to play a role in political developments in Sri Lanka in the late 1990s and early 2000s. Dhananjayan Sriskandarajah (2005) lays out the role of the internet in creating a transnational Sri Lankan Tamil discursive space (also see Cheran 2001). While coverage of the war was minimal in the international news media, “new media have created new spaces in which the diaspora can communicate within itself and the rest of the world in an effective, low-cost manner” (Sriskandarajah 2005, 499). Websites such as www.tamilnet.com and www.tamilcanadian provided members of the diaspora news and commentary from a Tamil perspective, while www.sangam.org and www.tamilnation.org (now offline — back online – Ed.) functioned as important sites for political discourse. Even pro-LTTE sites like www.eelam.com (now offline) were registered outside Sri Lanka. Sriskandarajah addresses the widely circulating critique that the Sri Lankan Tamil diaspora supported the LTTE and its secessionist ambitions when it was physically removed from the consequences of those actions. He points out that this claim overlooks the diversity of political views among diaspora Sri Lankans as well as the dialectical relationship that existed between the diaspora, the LTTE, and the Tamil homeland population (2005).

Tamilnet.com played a big part in shaping Sri Lankan Tamil political activism. Founded in 1995 by an eastern Tamil activist and writer in collaboration with expatriate engineers and computer scientists living in North America and Europe, it initially failed due to lack of access to on-the-ground information. Its founders quickly resolved that problem by creating a transnational news network that involved paying, training, and equipping Sri Lankan reporters. The website focused on stories from the north and east that were often overlooked by the Colombo-based mainstream press. The stories were written in Tamil, then translated into English in Colombo, and uploaded to the site in both languages. This system allowed the site to circumvent the biases of the Colombo press and avoid the Sri Lankan government’s strict censorship of the media (Whitaker 2004). Tamilnet.com has been widely described in Colombo as an LTTE mouthpiece, but Mark Whitaker writes that this claim is at odds with the neutral and accurate reporting. He also observes that while all of those involve in Tamilnet.com were Tamil nationalists, none of the editors or financial supporters of the website were members of the LTTE, and the LTTE had no financial stake in the website (2004).

The rapid growth of social media in the last decade has created new possibilities for transnational Tamil or Tamil-speaking political engagement. The semipublic nature of Facebook (posts may be viewed by “friends,” “friends of friends,” or the general public) means that individuals can engage in discussions and debates with people from a wide range of geographical, ethnic, religious, class, and sociolinguistic backgrounds. These interactions are unpredictable by nature since a person may post an image, text, or video meant for a particular group of people that may be viewed by other individuals and groups at different moments in time (Hillewaert 2015). Sri Lankan youth and young adults began widely using Facebook in 2008, and it is now the most popular social media network in the country (Thuseethan and Vasanthapriyan 2015). In 2009 second-generation diaspora Tamils, who had seen the LTTE’s struggle for an independent Tamil nation in the north and east (Eelam) fail, began advocating for Tamil rights in Sri Lanka on social media, a movement initially dubbed the “Blackberry Revolution” (Asokan 2009).

The Circulation of Tamil Signage Errors on Social Media

During my ethnographic research on the tension between ethnic conflict and multilingual education policies in the linguistic and social practices of Tamil and Muslim youth in Kandy and Colombo, my informants frequently drew my attention to the absence of Tamil on public signs or errors in the Tamil content. In 2008 an Up-Country Tamil university student took me on an autorickshaw ride through the center of Colombo in the course of which she pointed out multiple signage errors. Sri Lankans can report violations of the Official Languages Policy to the OLC, but it has no punitive powers. The NGO Centre for Policy Alternatives and the website citizenslanka.org track violations, which include the absence of Tamil or English from public signboards and documents (e.g., circulars, police reports, birth certificates, and licenses) and the dearth of Tamil interpreters at government offices and police stations (Wickrematunge 2016).

It was not until Arshad (the NGO volunteer whom I met in Kandy in 2007) messaged me in 2015 that I learned of the recent practice of circulating images of Tamil signboard blunders on Facebook, Tumblr, Twitter, WhatsApp, and other platforms. Although the types of errors differ, the images most widely circulated on Facebook involve the misspelling of Tamil words or phrases in ways that are offensive or derogatory. These examples are different from malapropisms—“the usually unintentionally humorous misuse or distortion of a word or phrase”—because their intentionality is indeterminate.¹⁰ Tamil speakers might comment on the blunders in English or Tamil, which is usually written in Tamil script.

The circulating images of Tamil signboard errors are numerous. In one image of a road sign, the Tamil word for city or town, naharam, is misspelled as naraham ‘hell’ (see fig. 2). In another example, the Tamil word for Kandy (kaNDi) on a government bus sign is written in Tamil as kuNDi ‘butt,’ or ‘bottom’ (see fig. 3). Last year an Up-Country Tamil Christian friend who works in a government office in Colombo shared a photo of a government document on her Facebook wall and several Facebook groups. Instead of “zonal education office” (kooDa kalvi kaariyaalayam), the Tamil portion reads “bad education office” (keTTa kalvi kaariyaalayam) (see fig. 4). Several images of Tamil errors in trilingual bathroom signs have also been in circulation. In one example from an unknown source, the Sinhala (kaantaa) and English say “ladies,” but the Tamil says “gents” (aaN) (see fig. 5). This error draws immediate attention to the indexical relationship between the signboard and its object (the bathroom), since the mistake can result in a person walking into the wrong gender toilet.

The materiality of the signboard images enables their effective circulation on social media. Facebook posts that include images or short videos are much more likely to get high engagement in the form of “likes,” comments, and shares than text-based posts (see Varis and Blommaert 2015). “Tamizh mozhiyai amulpaDuttu” (Tamil Language Implementation) is a public Tamil-language Facebook group started in 2015 by Tamils living in Colombo and the north and east. (It has a modest 1,594 members.) Its “About” page states in Tamil that since successive governments have failed to implement the co-official language policy, its mission is to “help ordinary people achieve their language rights.” The aim of the group is thus broad, but most of the posts include images of Tamil errors on trilingual public signboards and government documents, which can be independently searched with the “photos” option. It is clear from the overall content of the posts that its administrators do not view signage as the most significant aspect of the nation’s trilingual policies. Rather, the images of signboard blunders help attract attention to the group by emblematically presenting the dearth of language rights for Tamil speakers in postwar Sri Lanka. It is also arguably more visually effective to post images of trilingual signs with Tamil errors than Sinhala-only signs.

Interest in circulating images of language errors in Sri Lanka is certainly not limited to Tamil. “Typo in Colombo” and “Sri Lankan Spelling Mistakes” (SLSM) are Facebook pages (formerly on Tumblr) devoted to English errors on signs, advertisements, documents, websites, menus, and so on. These pages also occasionally post images of Tamil and Sinhala signboard errors. The SLSM administrators are a group of young middle-class Sinhala men from Colombo who attended English-medium private schools (most Sri Lankans study in Sinhala or Tamil). They faced controversy in 2016 when they posted an image of a sign for the University of Peradeniya Faculty of Arts that read “Faculty O Farts” (see fig. 6). Several people commented that the image, which was likely digitally altered, was derogatory to the university. This prompted the administrators to post a disclaimer expressing their humorous and lighthearted intention in posting the image. Some online critics argued that pointing out English typos in Sri Lanka underscores inequalities in access to English related to class and socioeconomic level. In the United States and European countries there is particular interest in circulating images of humorous typos or awkward phrases from East Asian countries, including Japan, Korea, and China, where English has no official administrative role but is widely learned as a foreign language (Kachru and Nelson 1996). The circulation of Tamil typos is distinct from English typos in the Sri Lankan context because of their deep political salience in the postwar moment.

Interpreting “Reserved for Pregnant Dogs”

Naay ‘dog’ is a severe pejorative term in Tamil that can be used for both men and women.¹¹ It is not equivalent to the English term bitch, although a person wrote in the Colombo Telegraph comments that the error likened Tamil women to “pregnant bitches.” This error is particularly salient because of the deep historical association in Tamil South Asia between the Tamil language, the Tamil people, and a mother. For Dravidian nationalists in South India in the mid–twentieth century the goddess tamizhtaay ‘Mother Tamil’ became a “focal sign for rallying the Tamil people against imperialism from abroad and from North India” (Bate 2009, 169). The discourses of Tamil devotees in the late nineteenth and twentieth centuries often contain slippage between tamizh ‘Tamil’, tamizhtaay ‘Mother Tamil’, taay paal ‘mother’s milk’, taay ‘mother’, and taay mozhi ‘mother tongue’ (see Ramaswamy 1997, 17). Tamil nationalist movements in Sri Lanka differed from those in India because they were driven by Sinhala/Tamil ethnonationalism. A reference to Mother Tamil in a 1977 Sri Lankan pamphlet, in fact, promoted the need to fight Sinhala domination (1997). In the following I analyze how three Sri Lankan Tamil speakers—Arshad (Muslim living in Belgium), Ruban (northern Tamil Christian living in the United States), and Darshini (northern Tamil Christian living in Sri Lanka)—drew on disparate semiotic ideologies to interpret the significance of the “reserved for pregnant dogs” blunder.¹²

Arshad: Intentionality versus Social Effect

Arshad is a 30-year-old Sri Lankan Muslim man from a small Muslim village outside Kandy. When we met in 2007 he was volunteering for an international NGO devoted to promoting peace and intercultural understanding. In 2010 he emigrated to Belgium. When he sent me the link for the Colombo Telegraph article in 2015, he wrote that the government still has a lot of work to do to properly implement their language policies. He told me he was also sending it to his Sinhala friends “to show [them] how important it is to know about this case.”

Arshad’s first language is Tamil (he studied in the Tamil medium), but he is also highly proficient in Sinhala and English. Like most southern Muslims, he frequently pointed out his detachment from the Tamil language as consistent with his religion-based ethnic identity (see Davis 2018). Despite this claim, however, he often tried to get the Sinhala NGO volunteers to take an interest in speaking Tamil, in line with the trilingual aims of the NGO. He also frequently sang classic Tamil film songs, patiently translating the lyrics into Sinhala and English for the Sinhala and foreign volunteers.

When I asked Arshad if the error had been made on purpose, he said “no.” He added that it was likely translated from Sinhala by a Sinhala person “who studied a bit of Tamil and got a job in a government firm with the push of a minister.” He added that half the money that goes into the trilingual programs often ends up in the pockets of corrupt politicians. He brought up inequalities in access to employment in Sri Lanka by noting that many Muslims and Tamils in Kandy are bilingual in Tamil and Sinhala but are overlooked for government translation jobs.

Arshad mentioned that I should read the Colombo Telegraph comments because someone there referred to the “reserved for pregnant dogs” error as “Tamil language genocide” (see above).¹³ He then added that a Tamil friend from the city of Kalmunai in eastern Sri Lanka had called the error tamizh kolai ‘Tamil murder’ in a Facebook message. This phrase, which was a hashtag in several tweets about Tamil signage blunders, is commonly used in reference to the incorrect use of Tamil and does not necessarily imply genocide.¹⁴ Still, it is sometimes likened to terms denoting genocide, which have long been used to describe the actions of the Sri Lankan state toward Tamil people (Varatharajah 2015). The UN Ad Hoc Committee on Genocide, which met between April and May 1948, included a definition of cultural genocide that was subsequently removed: “any deliberate act committed with intent to destroy the language, religion, or culture of a national, racial, or religious group on grounds of national or racial origin or religious belief of its members” (United Nations Economic and Social Council 1948, 18; see Skutnabb-Kangas 2000; Frieze 2012). As Sinthujan Varatharajah (2015) discusses, by the 1950s and 1960s some Sri Lankan Tamils, presumably influenced by the Ad Hoc Committee’s definition of genocide, began to discuss the violence against Tamils as an effort to destroy the Tamil nation and its cultural foundations. Over the next few years Tamil political parties and organizations popularized Tamil terms denoting genocide: ina paDukolai ‘racial massacre’ and ina azhippu ‘racial destruction’.¹⁵ In 1976 the Tamil United Liberation Front used the term cultural genocide in the Vaddukoddai Resolution (published in Tamil and English), which laid out the aspiration for an independent Eelam in the north and east of the island (Varatharajah 2015).

Following the anti-Tamil riots of July 1983, the focus of discussion shifted from the destruction of the Tamil language and culture to the destruction of the Tamil people. During the war years it was common for Tamil writers to conceptualize the state’s violence against them as racial massacre or destruction, but similar language was rarely used in English. However, during the army’s campaign to defeat the LTTE in 2008 and 2009, Tamil diaspora activists started to use the language of genocide in English and other European languages to try to convince India and Western nations to intervene and stop the killing of Tamil civilians. This language spread into the Western news media. Since 2009, the conceptualization of the civil war as a physical and/or structural genocide has been widely discussed and debated both in Sri Lanka and diaspora communities (Varatharajah 2015; Walton 2015).¹⁶ In 2015, the Northern Provincial Council passed the Tamil Genocide Resolution of 2015, which sought a UN inquiry to investigate genocide against Tamils by successive governments. It stated, “The acts of physical, cultural, and linguistic violence against Tamils are tantamount to genocide under the mental harm standard because extensive destruction of the Tamil culture and language threatens the Tamil people’s survival on the island” (Northern Provincial Council 2015). Some recent news websites and social media discussions refer to incorrect or inappropriate uses of Tamil as acts of genocide. For example, an article on Tamilnet.com published in December 2015 cited the incorrectly transliterated Tamil on a banner made by a US Embassy program in Jaffna as linguistic genocide. It stated, “After the three decades interlude of Tamil militant struggle, nothing has changed, except for the fact that the USA is now leading Colombo’s outlook of linguistic genocide.”¹⁷

Contextualizing the genocide and tamizh kolai comments, Arshad, referring to Tamils and not Muslims, said that the controversy around the “reserved for pregnant dogs” blunder is “about how much someone will like their language. It’s an identity for some people.” He added that people are still recovering from the “wounds of war, which makes everything more painful.” He said he wished he could post the “reserved for pregnant dogs” image so he could see how people would react, but he tries to stay out of politics on Facebook. In our face-to-face conversations Arshad often responded to news of injustices toward Tamil speakers with laughter, a reaction he replicates in Facebook interactions with laughing emojis. Trained in diplomacy by the NGO leaders, he was careful to give a balanced and nuanced perspective whenever I asked him to comment on particular situations. The language genocide comment in the Colombo Telegraph article implied that the blunder was part of the government’s systematic attempt to destroy the Tamil language. Arshad expressed a different semiotic ideology regarding the intentionality of the blunder by attributing it to a Sinhala government employee. But while he was happy to discuss the intentionality of the language error when I brought it up, he seemed more concerned with its social effects. His interest in the blunder is consistent with the well-known point in linguistic anthropology that interpretation is not contingent on intention (Duranti 1993; Gershon 2010). Arshad saw the error as disturbing to Tamils because of their language-based identity, but as a first-language Tamil speaker he is also clearly invested in Tamil language rights in Sri Lanka.

Ruban: “Funny and Sad at the Same Time”

Ruban is a Tamil Christian male in his early thirties who came to the United States a decade ago to pursue a BA degree in engineering. He grew up in Mannar in northern Sri Lanka, where he studied in Tamil-medium schools. He is part of a prominent academic family and has relatives in Sri Lanka, the United States, and elsewhere.

After hearing about my interest in trilingual signs, in 2015 a northern Sri Lankan Tamil Hindu colleague born in Toronto tagged me in a Facebook post soliciting images of Tamil signage errors. Within three minutes, Ruban had posted five images of erroneous Tamil signs. He added fifteen more over the next few days, including the “reserved for pregnant dogs” image. When I contacted him directly over Facebook Messenger, he told me that many of his Tamil friends collect images of Tamil signage errors on their phones and share them on Tumblr and Facebook. He mentioned that one friend has a personal database of hundreds of images.

When I asked Ruban about the “reserved for pregnant dogs” sign, he said it was “funny and sad at the same time.” A Tamil Hindu woman from Tamil Nadu made a very similar comment in Tamil on a public Facebook post on the blunder, noting that when she saw the typo she experienced a mix of laughter (sirippu) and sadness (varuttam). Ruban noted that the unintentional replacement of “ta” with “na” was derogatory. When I mentioned Arshad’s friend’s tamizh kolai comment, he wrote, “Spelling mistakes in general bother me whether it be Tamil or English. I took Tamil very seriously in middle school. I also find that these mistakes irritate Tamils in general—make them feel less important.” Ruban pointed to the humorous and disturbing aspects of the error, namely, that what we find funny can also be unsettling. He likely mentioned being bothered by typos in general to deflect my mention of the “Tamil murder” comment. But while he shifted the conversation away from explicit politics to the importance of correctness in language (also political, of course), he did not fully dismiss the significance of the blunder but admitted that errors are demeaning to Tamil people.

Collecting, posting, and commenting on images of Tamil signboard blunders give Ruban and other Tamil speakers living in-country or abroad a chance to perform their multilingual expertise. This is importantly related to their individual life histories: Ruban’s full fluency in spoken and written Tamil is a product of his primary and secondary education in Sri Lanka (he likely spoke English at home). Participants in public discussions on Tamil errors on Facebook and other platforms assume some degree of shared knowledge of Tamil, whether they comment in Tamil or English or leave emojis. I noticed that when individuals ask for clarification on the meaning of a Tamil word or phrase in a Facebook discussion, it is often read as a signal of their relative outsider status to the group. For example, several Tamils and Muslims responded to a post on the Kandy bus sign (see fig. 3) on SLSM’s Facebook page (people from all ethnic and religious backgrounds “like” the page) with “lol” or left laughing emojis. However, when a Sinhala man asked what kuNDi means, a woman with a Sinhala name who seemed to know Tamil, wrote that he was “lame” and “should grow up.” After the Sinhala man pleaded, “Just tell us what it says on the bus,” a Tamil man responded that it means “‘ass’ in Tamil.” Thus, Facebook discussions on Tamil blunders function as sites for the negotiation of emergent, fragile transnational Tamil-speaking publics.

While Ruban and Arshad focused on the social effects of public Tamil language blunders, other Tamil speakers explicitly debated their intentionality in the Colombo Telegraph comments (see above) and on Facebook. A northern Sri Lankan Tamil Hindu residing in Switzerland expressed a common view in stating in the Facebook group that the substitution of “na” for “ta” in the “reserved for pregnant dogs” blunder is an easy mistake. She added, “Even Tamil people here don’t know how to write Tamil so how can we complain about them [Sinhalas]?” A Tamil Christian man from eastern Sri Lanka, however, commented in Tamil that the mistake was obviously done wantonly because the meaning is so abusive. He noted that the error is similar to the writing of seevai ‘service’ as veesai ‘prostitute’ in another trilingual sign. A northern Sri Lankan Tamil Hindu woman living in Canada also assumed that the error was malicious in a comment in Tamil on a public Facebook post. She stated, “You guys [the Sri Lankan government] are not only killing the Tamil language but also pregnant women.” Arshad and Ruban both argued that Tamil signboard errors are insulting and demeaning to the Tamil people, but Darshini employed an entirely different semiotic ideology in questioning the very relationship between the posted signboard images and the on-the-ground situation in Sri Lanka.

Darshini: The Study of Signboard Errors Is Problematic

Darshini is a Jaffna Tamil Christian academic and activist who lives in Kandy but often travels abroad. I first met her on Facebook in 2015 when she responded to my Toronto Tamil Hindu friend’s Facebook post soliciting signboard errors. In response to the images Ruban shared, she wrote that the study of Tamil signage errors is “problematic for multiple reasons.” She elaborated in a second comment that when you look at typos you “incorrectly privilege linguistic form, when there should be more attention paid to the availability of services in particular languages in places like hospitals and police stations.” She said language studies of this nature “decontextualize language” and added that they also “reformulate politics as one about correctness about language and not about language use and people.” Further explaining her point, she added that a focus on typos “reinforces dominant thinking, trivializes the issue, and raises the question of whom we are representing or even talking to.” She added that people should focus not on typos, which she called a “diaspora occupation,” but rather on the absence of particular languages from public signs.

I sent Darshini a private Facebook message telling her about my academic background and research interests. When I said that my interest in Tamil signboard errors was part of a larger project on trilingual politics in postwar Sri Lanka, she expressed support for my research. I got a chance to meet her at a major university in the United States a few months later, but we mainly discussed our mutual friends in Kandy.

Darshini initially commented that a focus on Tamil signage errors privileges linguistic form over attention to “language use and people.” But, of course, trilingual public signs in Sri Lanka cannot be separated from lived social practices. As Arshad and others suggested, the “reserved for pregnant dogs” blunder was likely made by a Sinhala government employee who was not fully proficient in Tamil. Errors on public signs and in government documents reflect problems in the hiring of translators, as well as insufficient care around proofreading. Thus, while some Tamil speakers may treat Tamil typos in government signs as emblematic of the dearth of language rights for Tamil speakers, they are just one component of a larger problem.

If I understand Darshini’s comments correctly, I think she was arguing that the act of circulating images of Tamil signboard errors on social media removes the signboards from their original physical context (a bus, street, or government building). I suspect she is not concerned simply with the fetishization of correct Tamil spelling and grammar but with its overt politicization in postwar discourse. Darshini made the astute point that while images of errors circulate on social media, the less remarked absence of certain languages from signs is actually more problematic for people living in Sri Lanka. For example, in 2007 I noticed a Sinhala-only sign that marked a section of a major Colombo road as one way, a problem that could lead to traffic accidents (see Wickrematunge 2016). Darshini also stressed that people should focus not only on errors and omissions but also on the availability of services. The dearth of Tamil-speaking staff at hospitals, police stations, and government offices in the south causes severe hardships for Tamil speakers who are not competent in Sinhala.

Darshini’s early contention that interest in posting and commenting on Tamil signage errors seemed a “diaspora preoccupation” relates to a later statement about how discussion of decontextualized sign images brings up issues of audience or reception. She argued that it is mainly diaspora Tamils—a group often falsely assumed to be uniformly pro-LTTE (Amarasingam 2015)—who are circulating and commenting on Tamil blunders when they themselves are not directly impacted by trilingual policy failures. This point invokes the pervasive critique of diaspora Tamils for financially supporting the LTTE and its struggle for Eelam from a position removed from its consequences (Sriskandarajah 2005). It is thus clear that Darshini relates the concern with Tamil blunders to Tamil nationalist and diaspora-centered perspectives.

Contrary to Darshini’s view, I found there to be an equal number of resident and diasporic Tamil speakers circulating and commenting on Tamil errors on Facebook. In addition, my Tamil-speaking friends living in Sri Lanka were just as interested in discussing language errors as those in diaspora communities. As a reflection of demographics, resident Sri Lankans who commented on blunders included North and East Tamils, Up-Country Tamils, and Muslims, while the diasporic Sri Lankans were mainly North and East Tamils.¹⁸ In 2018 I asked two Sri Lankan Tamil Hindu academics who grew up in Toronto if they thought the interest in Tamil errors in public signs and documents is indeed a diaspora preoccupation. One responded, “Not necessarily.” He then discussed the circulation of signboard errors as social action, stating, “While strands of the diaspora are preoccupied with sign images, I would say it is about justice and accountability for what happened in 2009.” The colleague who initially tagged me on the Facebook post soliciting Tamil signage errors said the issue is not about “diaspora/non-diaspora but silly/non-silly.” He explained, “It’s silly to post images of errors and laugh,” but they should still be posted to “highlight how language continues to be such a sticking point in the conflict, and why majoritarian indifference often leads to these mistakes.” My colleague is thus against posting images of Tamil signage errors in a gratuitous manner, but, similar to how they are used on the Tamil Language Implementation Facebook group (administered by Tamils living in Sri Lanka), they can point to the plight of Sri Lankan Tamil-speaking minorities. Thus, while Darshini disapproved of the widespread circulation of blunders on social media, many others in the diaspora and in Sri Lanka view it as a necessary and useful political strategy.

Chandrika, a Sinhala-Buddhist NGO worker whom I met in Kandy in 2008, agreed with some of Darshini’s views. In 2018 a mutual Up-Country Tamil Christian friend named Vijay, who had come to the United States for graduate school in 2007, commented on an image of an erroneous trilingual bathroom sign posted on Facebook (see fig. 5), noting that it is an example of “failed reconciliation” and the “government’s disregard for Tamil people.” Chandrika wrote that Vijay was ignorant in thinking that a “simple language mistake means there’s no place for Tamils in Sri Lanka.” When I contacted Chandrika privately, he noted that it is common for “diaspora Tamils or even diaspora Sinhalas” to post controversial comments on Facebook without understanding the actual situation in Sri Lanka. He cited the Sri Lankan Tamil recording artist, M. I. A. (Mathangi Arulpragasam), as an example of a diasporic Tamil who misleads the world with her biased statements. He said he was a fan of hers until he saw her criticize President Sirisena for not including Tamil people in the reconciliation process in a January 2015 Channel 4 interview. M. I. A., whose father has ties to the Tamil separatist cause, lived in northern Sri Lanka as a child before immigrating to the U.K and represents herself as a voice for Sri Lankan Tamil people in her music videos and interviews (Fuller 2009).

Disparate Semiotic Ideologies and Their Political Implications

Tamil speakers’ interpretations of Tamil signage errors reflect contrasting semiotic ideologies concerning the intentionality of the errors and the relationship between the posted images of the errors and on-the-ground sociolinguistic practices (Keane 2003, 2018). Their views are informed by broader sociocultural and political perspectives, some of which circulate within ethnic, religious, or geographic groups. The opinion that Tamil blunders are deliberate acts (either committed by individuals or the government as a whole) is a minority view. Almost all of the people who made these comments in Facebook discussions were North and East Tamils (Hindu or Christian) residing in Sri Lanka or in the diaspora.¹⁹ This finding is not surprising given the severe animosity and distrust that many Tamils from this region feel toward the Sri Lankan state following the devastating events of the war and its aftermath. In addition, as I discussed above, the war is widely framed as genocide in the north and east and in the diaspora (Varatharajah 2015). But it is also notable that not all who view the war as a physical and/or structural genocide against Tamils would consider a Tamil blunder on a public sign a genocidal, or even malicious, act.

My three main research participants—Arshad (Muslim living in Belgium), Ruban (northern Tamil living in the United States), and Darshini (northern Tamil living in Sri Lanka)—subscribed to the common view among Sri Lankan Tamil speakers (North and East Tamils, Up-Country Tamils, and Muslims) that language blunders are unintentional. Arshad, who provided the richest observations on this matter, attributed responsibility for the “reserved for pregnant dogs” error to a single Sinhala government worker. But he, like many others I spoke with, saw Tamil errors on public signs and government documents as stemming from government corruption and systemic inequalities. In a conversation over Facebook Messenger in 2019, Arshad noted that people’s views on the ethnic conflict shape how they interpret the Tamil errors. While the opinion that such blunders represent affronts, insults, or even genocidal acts leaves little hope that the government will incorporate Tamils or Tamil-speakers into the nation, Arshad and others still see some possibility for a sustainable future.

Mano Ganesan, the Tamil Hindu minister of National Co-existence Dialogue and Official Languages from 2015 to 2018, has acknowledged Tamil and other signage errors on his Facebook page and in press interviews. In a 2017 interview with The Economist he referred to the “reserved for pregnant dogs” blunder as an embarrassing mistake and admitted that a fully bilingual (Sinhala and Tamil) bureaucracy is decades away. His comments help to alleviate controversies over Tamil blunders by casting them as a part of the government’s ongoing efforts to create a trilingual nation (see Davis 2020). In the same interview he stated that the full implementation of the Official Languages Policy “will be a prelude to a political solution,” one that will address Tamil grievances.²⁰

Darshini’s opinions on language errors were distinct from most others in that she critiqued the very focus (by others and by me) on images of Tamil signage errors posted on social media. She called into question the necessary relationship between a sign (those posted images) and its object (lived sociolinguistic practices) (Keane 2003). While the Sinhala NGO worker from Kandy agreed with some of her views, my two northern Tamil Hindu colleagues from Toronto countered her claim that interest in Tamil blunders is a diaspora preoccupation by arguing that images of Tamil signage blunders underscore the indifference of the government toward Tamil speakers (my Up-Country Tamil friend Vijay also expressed this view). Thus, while Darshini suggested that the circulation of signboard images can draw attention away from efforts to ensure language rights for all Sri Lankans, my colleagues argued that they can reveal important realities. The disparate semiotic ideologies that Darshini and my Toronto colleagues drew on engage the timely question of which Tamils—those residing in-country or abroad—can and should create a future for Tamils in Sri Lanka.

After the LTTE’s 2009 defeat the idea that the diaspora would lead the struggle for Eelam was exemplified when the Provisional Transnational Government of Tamil Eelam, which identified with the LTTE, held elections in May 2010 in eleven different countries (Thiranagama 2014). However, the International Crisis Group (2010) reported that the diaspora’s continued support for the LTTE and the separatist cause had increased the gap between diasporic and resident Tamils, who were exhausted by decades of war and wanted to rebuild their lives in Sri Lanka. L. Vimalarajah and R. Cheran, echoing Sriskandarajah’s views (2005), argue that the relationship between the LTTE and the Tamil diaspora is “complex and multi-faceted” (2010, 7). They also suggest that the quest for independence cannot be automatically associated with the LTTE, since it predates the group and represents different things to different people. They critique the view that the diaspora does not have to face the consequences of its long distance nationalism as shortsighted; they say that perspective only “trivializes the pain and trauma of thousands of diaspora Tamils whose family members and relatives have perished in large numbers in the last few months of the war” (16).

Significant shifts in Tamil diaspora activism occurred after May 2009 (Thiranagama 2014). After many of the LTTE’s diaspora representatives withdrew from politics, a space opened up for second-generation diaspora Tamils, many of whom had been critical of the LTTE, to participate in the Tamil nationalist movement (Vimalarajah and Cheran 2010). These social media-savvy youth, whose politics have been shaped by their variegated relationships to Sri Lanka and their host countries, hold different views about how to resolve the Tamil question (Amarasingam 2015). Cheran stated in a recent interview that Tamils living in Sri Lanka should be the ones to decide whether or not to pursue a separate Tamil state, but because the Tamil nationalist movement is transnational it is not possible to discuss any project in Sri Lanka without taking the diaspora into account (Cheran and Halpé 2015). Capturing the possibilities and limitations of diaspora activism, he writes, “Right now, it seems to me that the fight at the literal, theoretical, and conceptual level is a fight between the irresponsibility of distance and the responsibility of distance. This is going to determine what sort of trajectory the demand for a separate state called Tamil Eelam will take in the future” (2015, 106).

Academics and activists in Sri Lanka and abroad have also stressed the need for the Sri Lankan government to consider the interests of all Tamil-speaking minorities (Thiranagama 2011; Cheran and Halpé 2015). As mentioned above, although most Muslims speak Tamil as a first language, their connection to Tamil is mediated by their religion-based ethnic identity (Davis 2018). While southern Muslims commonly stress their lack of attachment to Tamil, Tamil language rights constitute a key element of Muslims’ imagined political futures, particularly in the predominantly Tamil-speaking east and north of the island (McGilvray 2008).

Conclusion

In the last three decades the internet and social media have enabled the creation of a transnational Sri Lankan Tamil-speaking public. While digital forums bring together Sri Lankan Tamil speakers from different ethnic, religious, and regional backgrounds, they also reveal points of contention in the effort to create a common Tamil or Tamil-speaking postwar experience. Resident and diasporic Tamil speakers are both interested in circulating and discussing images of Tamil blunders on public signs and government documents. I have shown how their disparate semiotic ideologies—regarding the intentionality of language errors and the ability of circulated media images to accurately represent the world—have implications for imaged postwar political possibilities as well as the question of whether effective Tamil activism can be carried out from a physical distance.

Scholars of digital culture have recently argued that it is now difficult to separate online and offline worlds (Varis 2015). While this is certainly an important point, we should not overlook the fact that in order to understand social media discourses we must explore how people understand circulated images, videos, and text in relation to their own experiences in, and perceptions of, the physical world. Keane’s (2003, 2018) concept of semiotic ideologies provides an excellent framework for thinking through the interplay of digitally and nondigitally mediated forms of signification. Its concern with the nature, function, and consequences of signs, as well as its emphasis on the openness of ideological processes to destabilization, suggest how we might understand just what processes occur across the increasingly permeable zone between digital and analog culture and politics. As my attempt here to delineate a “representational economy” for Tamil signage blunders has shown (Keane 2003, 41), Sri Lankan Tamil’s speakers’ understandings and interpretations of linguistic forms and images circulated online both reflect and produce their conceptions of the postwar future.