Article body

1. Introduction

Linguistic and translation justice have been of growing interest in translation studies recently (Meylaerts 2011; De Schutter 2017; McDonough Dolmaya 2017). Research in this area focuses on how to design and implement fair language and translation policies—the rules that regulate the use of language and translation in education, legal affairs, political institutions, the media, and administration (Meylaerts 2011: 744-745). Indeed, translation and language policies more broadly have received increased attention from translation studies researchers who explore questions related to how (non-)translation can help ensure equal opportunities for everyone in a society (Meylaerts 2011: 753). In some cases, researchers have focused more particularly on the links between minority languages and translation policies (Córdoba Serrano and Diaz Fouces 2018), arguing that dominant language groups have language rights that are not necessarily enjoyed by minority groups because the former are in a prominent position within a political regime while the latter are not. As Meylaerts asserts, it is not clear what kind of language and translation policy best encourages integration, participatory citizenship and social cohesion in multilingual societies (Meylaerts 2011: 753), all of which are important issues for translation studies researchers to explore.

One way for people to access information about local government services and activities and actively participate in their communities is through the websites developed and maintained by the municipalities in which they reside (Cullen, O’Connor, et al. 2003; Liste and Sørensen 2015; Dolson and Young 2012: 2). Providing public access to government information is crucial to good local governance, since it helps assure government accountability while also facilitating public participation in local government and enhancing public access to community services (Cullen, O’Connor, et al. 2003, 191). However, when members of the public are not proficient in the local official language(s), their ability to participate in their local communities and to access municipal services can be impacted. As Meylaerts (2011: 743) notes, language is crucial to establishing modern democratic societies, “especially in their search for cohesion through participatory citizenship.”

Municipal websites therefore offer an intriguing glimpse into how language and/or translation policies are implemented by local governments. They allow us to better understand who can access local information online and whether (and how) municipalities allow official and non-official language speakers to participate in their local communities. Numerous studies have examined the websites of municipalities in a wide range of countries, focusing on issues such as the effectiveness of the websites in conveying information to the public (see, for example, Cullen, O’Connor, et al. 2003), the content of municipal websites (Liste and Sørensen 2015; Holzer and Kim 2006; Dolson and Young 2012), and the priorities of municipal governments with respect to their websites (Liste and Sørensen 2015). While some work has explored language and translation (McDonald, Merwin, et al. 2011; Carroll 2010), many of the studies focusing on municipal websites consider the question of language only superficially, if at all. For instance, Dolson and Young (2012: 7), in their study of Canadian municipal websites, exclude Quebec municipalities “due to the French language barrier,” while Holzer and Kim (2006: 68) compare municipal websites in 100 countries using 98 criteria, including whether the websites offered “access in more than one language,” but they do not describe what exactly “access” meant (for instance, integrated machine translation engine, some pages in another language, a fully functional mirror site in another language).

And yet, the language of municipal websites is an essential element to consider. Governments cannot remain politically neutral with regard to languages, since they must communicate with their citizens and they must therefore choose one or more languages in which to do so (De Schutter 2017: 16): they do so via policies, a term that will be used in this paper to refer to the way a government pursues a particular goal (Grin 2003a: 21).[1] Policies can be explicit (that is, outlined in a written document) or implicit (in other words, inferable based on practice) (Meylaerts and González Núñez 2017: 2-3). These policies will determine the language skills people need to participate in political life, to access the labour market and to interact with others (Grin, Marácz, et al. 2014: 6) Thus, the language(s) in which municipal website content is available offers important insight into questions such as how a municipality has positioned itself with respect to its official/non-official language communities, how it views language rights for the various groups who live within its borders, and whether it feels public officials have a right or a duty to speak certain languages amongst themselves or with the public (Patten and Kymlicka 2003: 16-17). This article will attempt to contribute to the existing body of work on translation and language policy and translation and linguistic justice by comparing the support for official and non-official languages on the website.

2. Official and non-official languages in Canada

Canada is an ideal country for a case study, as it has two official languages at the federal level (English and French) and a significant percentage of immigrants—21.9% in the 2016 census (Statistics Canada 2017). While many of these immigrants speak English, French or both, only a quarter (27.5%) of immigrants in 2016 were native speakers of one of the official languages, and about 7% could speak neither English nor French (Statistics Canada 2017). In total, more than 170 non-official languages are spoken across the country—this figure includes more than 50 Indigenous languages and 120 new immigrant languages (Statistics Canada 2016b). These language categories are discussed in more detail below.

As noted by others (Lane-Mercier, Merkle et al. 2014: 471-472; Hebert 2016: 22), the fact that Canada has legislation related to official languages does not mean it also has an explicit translation policy: the Official Languages Act (Parliament of Canada 1969/2017), for instance, describes numerous contexts in which documents must be available in both French and English, implying that translation would be necessary; however, translation is directly mentioned only three times: when referring to the proceedings of Parliament, regulations by the Governor General, and information added to bilingual forms used in federal court (Parliament of Canada 1969/2017: 4, 6, 10). The situation is similar for non-official languages. Section 3(c) of the Canadian Multiculturalism Act (Parliament of Canada 1988/2014) stipulates that the Canadian government must endeavour to ensure that “individuals and communities of all origins” can help shape “all aspects of Canadian society” and that the federal government should help eliminate any barriers to this participation. Moreover, in Section 3(i), the Act specifies that the federal government should “preserve and enhance the use of languages other than French and English, while strengthening the status and use of the official languages of Canada.” However, neither translation nor interpreting are mentioned in the Act, and so it is not clear how the various levels of government should put these policies into practice.

Canada is, of course, officially bilingual, but only at the federal level. Meylaerts (2011: 752) suggests that Canada has adopted a policy of institutional monolingualism combined with institutional multilingualism—one where monolingualism prevails among local governments, while bilingualism is the policy of the federal government. Federally, Canada does have a policy of bilingualism, but the regional and municipal governments are not always monolingual. At a regional level, language policies differ widely from one province and territory to another. Only one province, New Brunswick, is officially bilingual (French/English). Quebec, under the Charter of the French Language, or Bill 101 (Assemblée nationale du Québec 1977/2020), is the only province to have French as an official language, while the other provinces are either officially or de facto English, although often with some official recognition of French in designated areas or contexts.[2] The three Canadian territories also recognize Indigenous languages in addition to English and French.[3] At a local level, several municipalities have explicit language policies officially recognizing—to varying degrees—both French and English. These include Moncton, in the province of New Brunswick (City of Moncton 2018), and, in the province of Ontario, both Ottawa (City of Ottawa 2015) and Sudbury (City of Greater Sudbury 2001).

While federal, regional and local language policy documents in Canada frequently focus on French and English, the country’s two official languages, fewer policy documents specifically address indigenous languages, and fewer still discuss the languages spoken by “new immigrants” (Meylaerts 2011: 744)—that is languages other than French, English or Indigenous languages like Cree and Inuktitut. This shortcoming, coupled with the lack of an explicit federal translation policy and the significant number of immigrants in Canada, presents challenges for municipalities where large numbers of immigrants settle. When municipalities seek to make information available to the public, they must choose in which languages to do so: one official language, both official languages, an official language and one or more non-official languages, etc. And, if the municipality chooses to make information available on its website in a non-official language, it must determine which languages to target and what information to disseminate.

To help investigate translation policies in the context of non-official languages in Canada, this paper will examine the municipal website of the city with the largest immigrant population: Toronto.

2.1. Demographic and linguistic profile of Toronto

More than half of Canada’s immigrant population lives in just three cities, Toronto, Montreal, and Vancouver, with Toronto having the largest percentage of its population (46%) comprised of immigrants (Statistics Canada 2017). Toronto is also one of the most linguistically diverse cities in the country: 44% of Torontonians have a mother tongue other than English and/or French, while slightly over a quarter of the city’s residents reported most often speaking a language other than English or French at home in the 2016 census (Statistics Canada 2016a). Indigenous languages are spoken by a very small percentage of the population: only 425 of Toronto’s 2.7 million residents reported having an Indigenous language as a mother tongue in 2016, and only 110 reported speaking one of these languages most often at home (Statistics Canada 2016a).

Because this paper will rely heavily on data from the Canadian census, labelling various language communities will not be straightforward. Statistics Canada offers data for both “mother tongue” (the language first learned at home in childhood, or, if a person no longer understands the first language, then the second language learned at home), and for “language spoken most often at home.” With the aim of striking a balance between concision and accuracy, this article will use the term official language speakers to refer to those who report speaking French or English most often at home (regardless of their mother tongue),[4]Indigenous language speakers to refer to those who report speaking an Aboriginal[5] language most often at home, and new immigrant language speakers[6] to refer to those who reported speaking a language other than French, English or an Indigenous language most often at home. When drawing on data that refers only to mother tongue speakers (and not to those who speak a language most often at home), this paper will specify that this is the case.

Slightly more than half of Toronto’s population has one of Canada’s two official languages as a mother tongue, with about 53% of residents being native speakers of English, and only 1% having French as a mother tongue (Statistics Canada 2016a). While more than 60% of Toronto residents speak English most often, about a quarter of the population speaks a non-official language most often at home, (Statistics Canada 2016a). Mandarin and Cantonese speakers are the largest non-official language group, comprising 24% of the non-official language speakers, while the next largest non-official language speaker groups, namely Tagalog, Tamil, Spanish, Portuguese, and Farsi, each comprise 4-6%.

2.2. Language and translation policies in Toronto

The City of Toronto has published several explicit policies related to translation into official, Indigenous and new immigrant languages. The Multilingual Information Provisions Policy (City of Toronto 2017a) directly discusses the circumstances under which material produced by the city should be translated, while By-law 1176-2009 “[t]o authorize the use of languages other than English in notices, forms (other than prescribed forms) and other information for regular municipal election and by-elections in the City of Toronto” (City of Toronto 2009) discusses the translation of material related to municipal elections. The provision of services in sign languages is not included in the Multilingual Information Provisions Policy, as this is supposed to fall under the city’s Multi-Year Accessibility Plan (City of Toronto 2011); however, the plan does not explicitly describe when and how ASL interpretation should be provided, other than noting that “communication supports” (which include ASL) must be provided upon request.

Translation has clearly been a subject on which the City of Toronto has been reflecting for some time. The Multilingual Information Provisions Policy was adopted by Toronto City Council on May 24, 2017, but it replaced the previous policy, the City of Toronto Multilingual Services Policy, which had come into effect in 2002 (City of Toronto 2002; see Hébert 2016 for a discussion of this previous policy). When preparing the 2017 policy, the City conducted a survey of Toronto residents to determine whether and how they accessed translated municipal information (City of Toronto 2017b), and also issued a report discussing the financial implications and rationale for the new policy (City of Toronto 2017c).

The City of Toronto does offer other kinds of translation and interpreting services (such as telephone interpreting in 180 languages for its 311 information service, and interpreting for public meetings). The city also translates signs that are physically displayed in the city (like development areas and construction sites) but are not available online. Some of these translation services are discussed in the City of Toronto Multilingual Information Provisions Policy, and others are discussed in related policies, such as the City of Toronto Multi-Year Accessibility Plan. However, because this article is exploring translation in the context of municipal websites, I will be discussing only translation and online texts.

The Multilingual Information Provisions Policy (City of Toronto 2017a) contains nine sections: the first three contextualize the policy and define key terms, while the last two discuss compliance and reporting, and list other relevant City of Toronto policies. The remaining sections pertain to the translation of critical information that could affect the health and safety of Toronto residents, material targeted at specific neighbourhoods, municipal marketing material, and all other kinds of information.

In the three contextualizing sections, several points are worth noting: first, the city’s multilingual information must comply with existing municipal policies, including the human rights policy, and the information cannot be discriminatory in nature. Second, machine translation must be offered as an option on the city’s website. Finally, although professional translation is mentioned and defined in the policy document, the definition is rather vague, specifying only that a “professional human translator” must perform the translation, but not what qualifications or training the translator should have.

Section 4 addresses the translation of critical information—that is information that could impact the health and safety of Toronto residents, and/or communications about disruptions of critical services such as public transportation or electricity. Among other things, it notes that critical information must be translated by a professional translator “into no less than the top ten languages spoken at home as indicated by the latest census data available, and French” (City of Toronto 2017a: 3), except when the information is targeted at a local community (in which case Section 5 applies), or when the information is targeted at a specific non-localized community, in which case the division that is preparing the information should conduct an assessment to determine which languages to target. In emergencies, information is to be disseminated in English and then professionally translated “as soon as possible” (City of Toronto 2017a: 4). Members of the public can also request information about critical services in any language, and the information must be translated “within a reasonable timeframe” (City of Toronto 2017a: 4).[7]

Section 5 focuses on information targeting specific Toronto areas. It stipulates that when this information is not available on the city’s website (and could therefore be machine translated), it will be professionally translated “when at least 5% of the population in that [geographic area, neighbourhood or ward] speaks languages other than English at home as per the latest census data available” (City of Toronto 2017a: 4). When this standard is met, the information will be summarized, and the summary will be “professionally translated into at least the top three languages spoken at home” (City of Toronto 2017a: 4), unless the information is targeting just a specific neighbourhood, in which case the summary will be translated into only “the top language spoken at home” in that neighbourhood (City of Toronto 2017a: 4). In either case, though, the city can assess the need for translation and choose to translate into fewer or no languages (City of Toronto 2017a: 4).

Section 6 covers the translation of marketing material, stipulating that a translation of full or summarized information should be considered based on the target audience. Provisions exist for both international and domestic audiences: material for international audiences can be translated into the “top spoken language at home” in that region, while material for domestic audiences can be translated into the official language(s) of that province, and/or the top three languages spoken at home in that municipality. A special provision addresses the translation of material that promotes Canadian history or cultural events: Section 6.1.3 notes that in such cases, the information “may be translated into French” (City of Toronto 2017a: 5).

Finally, Section 7 states that all other kinds of information will be translated on a case-by-case basis, as determined by the head of the division based on factors such as whether the translation will help encourage certain groups or neighbourhoods to participate in city programs or comply with and use city services, whether the translated information will be relevant for at least six months, and whether there is a budget for the translation.

Several points are of interest here: first, the City of Toronto’s policy document relies on the “language spoken most often at home” census data, rather than mother tongue data: Sections 4, 6, and 7 of the Multilingual Information Provisions Policy prioritize languages based on the number of speakers (for example, the top three languages spoken at home in a given district) while Section 5 requires a minimum threshold to be met (for instance, at least 5% of the population in a neighbourhood speaks a language other than English at home). Thus, some mother tongues may not qualify as target languages if they are not often spoken at home by Toronto residents. Yet, as Grin, Marácz, et al. (2014: 10) argue, the linguistic profile of many people includes several languages, which they draw on not only for communication, but also for identity building. Indeed, more than 200,000 Torontonians reported speaking English and a non-official language at home most often in the 2016 census, while more than 300,000 reported speaking a non-official language “regularly” (but not most often) at home (Statistics Canada 2016a). These responses are not broken down by language, however, so under its current policies, the City of Toronto could not consider these census responses when deciding which target languages to choose.

Second, the city does not treat all languages equally: in some cases, official, Indigenous, and new immigrant languages must all meet the same criteria to qualify for translation, while in others, an official language (namely French) is given priority. This is not unexpected since, as Grin (2003b) notes, all multilingual states treat languages with inequality because new immigrant languages are given less recognition.[8] French is spoken at home by less than 1% of Torontonians, and the same is true for all Indigenous languages combined. Thus, whenever the policy prioritizes translation into one of the top languages spoken at home, without having an explicit exception for French or an Indigenous language, these languages will not meet the requirement for translation. Notably, Sections 4 and 6 explicitly include French as a possible target language, but no sections offer similar exceptions for any Indigenous languages. This policy therefore exhibits both “translation as accommodation” (Córdoba Serrano and Diaz Fouces 2018: 6-10), for new immigrant and Indigenous languages (and occasionally for French) and “translation as a right” (Córdoba Serrano and Diaz Fouces 2018: 6-10), for French: this language is given special recognition because French and English speakers have certain rights both provincially and federally that speakers of other languages do not.

Bylaw 1176-2009 is shorter than the Multilingual Information Policy and covers only translation situations related to municipal elections. It stipulates that the city may choose to translate election information into languages “spoken and understood in the home” by at least 2% of a ward’s population at the time of the census (City of Toronto 2009: 1). Note, however, that the bylaw does not require the city to translate election material, since it says only that translations “may be prepared” (rather than must be prepared). And, like the previous policy document, the bylaw relies on the “language most often spoken at home” census data.

2.3. Translation justice

Before turning to the website to see how the city’s translation policies are implemented, I would like to discuss the concept that will underpin the theoretical discussion in the rest of this paper: translation justice.

A relevant framework for discussing translation justice is proposed by De Schutter (2017), who argues that translation justice involves choosing between different translation policies. He suggests that states should choose which language(s) will be “singled out for state-supported translation” (De Schutter 2017: 17) based on the underlying language goal—namely whether the translation policy aims to accommodate identity interests in a language, non-identity interests or a combination of the two. A policy that aims to accommodate identity interests is based on the principles of autonomy and dignity for target language speakers: under such a policy, individual identity, which is closely tied to one’s language, must be politically respected and secured though translation rather than by encouraging target language speakers to learn the majority language (De Schutter 2017: 21-22). A policy that aims to accommodate non-identity interests is based on the principles of efficient communication, democracy and equality of opportunity. Under such a policy, translation is intended to help target language speakers understand the source text message rather than to affirm the identity of target language speakers (De Schutter 2017: 17-18). Thus, translation is considered a temporary means of supporting communication and allowing target language speakers to participate in public life, but target language acquisition is the ultimate goal, since communication is most efficient when translation is not required (De Schutter 2017: 29). It is also possible for a translation policy to be based on a dual position founded on both identity and non-identity interests. Such a policy would involve full rights to translation in certain contexts (for instance, when exercising one’s right to vote, participating in the public sphere or communicating with government officials) but minimal translation rights, and promotion of language learning, in other contexts (De Schutter 2017: 28-29).

We could argue that non-identity (or instrumental) approaches promote assimilation, since they rest on the assumption that everyone should learn at least one of the dominant or official languages. In fact, De Schutter (2017: 20) notes that instrumental interests will most likely lead to “inculcating citizens with a shared language, because doing so is most effective, erases hurdles in democratic exchanges and, as such, is beneficial to finding job opportunities.” This is problematic, of course, if we want to preserve linguistic diversity in a region—something Grin (2003a: 26) suggests has a positive effect on the welfare of a society. However, encouraging people to learn an official language does not mean that they cannot also maintain their non-official languages: assimilation need not be the ultimate goal. Indeed, as Grin, Marácz, et al. (2014: 10) suggest, all multilingual citizens can develop close connections with the languages they have learned (whether formally or not) and incorporate these languages into their “communication repertoire” as well as their personal identity.

Instrumental interests are integral to the City of Toronto’s translation policy. Consider the opening paragraph of Section 1 of the City of Toronto Multilingual Information Provisions Policy:

The City of Toronto recognizes that providing information about its programs, services, and engagement activities in languages that reflect the linguistic diversity of the city is an important way for the public to engage, participate, and be informed of City services and programs.

City of Toronto 2017a: 1

Likewise, see the preamble to By-law 1176-2009:

WHEREAS Council [of the City of Toronto] wishes to allow election information to be made available to electors in any language which the most recent and available Statistics Canada data shows was spoken and understood in the home by at least two percent of a ward’s population at the time of the census

City of Toronto 2009: 1, emphasis added

In these two paragraphs, which contextualize two of the city’s translation policy documents, the city focuses on the instrumental interests highlighted by De Schutter (2017), namely efficient communication, democracy, and equality of opportunity, rather than affirming the identity of residents who speak a language other than English at home. The city wants to provide information about programs, services, and elections (communication), thereby encouraging participation in programs, services, and the electoral process (democracy and equality of opportunity).

Communication efficiency is evident in several aspects of the city’s translation policy and its translation practices. By basing most translation decisions on the “language spoken most often at home” census data, rather than the “mother tongue” data, the city positions translation as a way to temporarily communicate with residents who have not yet switched to speaking English most often at home rather than as a way to affirm the identity of those who once preferred to (and who may still also) speak a language other than English. Moreover, in its language services survey (City of Toronto 2017b), the city asked respondents about the factors the city should consider when determining whether to translate a document. In its multiple-choice options, the survey offered pragmatic factors—translation cost, importance of the information to be translated, number of translation requests, etc. – rather than identity-related factors such as whether translations should be produced to help preserve a language or to encourage its use in the city, indicating that such factors were not a priority when the Multilingual Information Provisions Policy was prepared. To see whether non-identity (instrumental) interests are prioritized when the policy is implemented, let us now turn to the example of the city’s official website.

2.4. The municipal website

2.4.1. Methodology

To explore how the City of Toronto’s translation policies are put into practice, the city’s website ( was searched for documents and pages in languages other than English. The homepage is available only in English, and although a hyperlink at the bottom of the page is labelled “Translate,” it merely takes users to a page with a drop-down box where they can select one of 51 languages into which they can translate the website using an integrated Google Translate tool.[9] The “Translate” link does not direct users to a list of documents or pages that have been professionally translated specifically for Toronto residents. Indeed, there is no specific page with a list of all translated documents/pages and the languages in which they are available. This makes accessing translated material via the city’s website rather challenging.

For this project, the search bar on the homepage was used to query the name of every language spoken most often at home by at least 15,000 Torontonians in the 2016 census (0.5% of the population). Queries were repeated using the native name of the language (such as français), its English name (French), and any other potential variants (like Filipino, Philipino, Tagalog).[10] Although they are spoken at home by very few Toronto residents,[11] the names of the Indigenous languages most often reported as a mother tongue of Toronto residents (that is, Cree, Ojibway, Mi’kmaq, and Inuktitut) were also queried. Whenever a translation was found in a language that was not initially on the list (for example, Pashto, spoken most often at home by approximately 3000 Torontonians), a separate query was done to see whether any other content was available in that language. In all, 58 queries were conducted in February 2019.

While this methodology is not perfect, since some languages may have been overlooked and some documents may not have been labelled with the name of the language in which they were written, it was chosen for two reasons. First, in the Multilingual Information Provisions Policy, the City of Toronto prioritizes translations based on the number of people who speak a language at home; adding queries for languages spoken by few Toronto residents was unlikely to have yielded more translations but would have greatly increased the research time.[12] Second, this process was the one that a typical Toronto resident would need to use if they wanted to find translated materials on the website: they are unlikely to write a script to search the website for content in a particular language. They could certainly come across translated material by chance when browsing through the website, but they would need to use the search engine if they wanted to systematically search for content in a language. Therefore, if content was not found for this project via the search engine queries, it is unlikely to be found by city residents either.

2.4.2. Overview of the translated content

Although the City of Toronto’s website is not “localized” in the sense of being adapted for other specific sociocultural regions (Jiménez-Crespo 2013: 12), the website localization levels proposed by Jiménez-Crespo (2013) still provide a helpful way to categorize the kind of translated content that is available on the website. Jiménez-Crespo’s model ranges from level 0, in which a website offers translated documents or machine translation links, to Level 4, in which a fully localized “mirror” site exists in another language. Given the low level of localization on the website, I have slightly adapted his model to distinguish between a site that offers the possibility of machine translation (Level 0) and a site that offers translated documents (Level 1), resulting in one additional level, as seen in Table 1:

Table 1

Levels of translation on the website

Levels of translation on the website
Categories slightly adapted from Jiménez-Crespo 2013

-> See the list of tables

As Table 1 illustrates, the City of Toronto’s website provides limited support for languages other than English. Its translation offerings consist mainly of translated documents, which Internet users can access only via webpages, search engine interfaces and navigation menus that remain in English. No Indigenous or new immigrant languages received support beyond Level 1. However, the website does offer support for non-Latin characters, as the embedded search engine returned results for languages in their native names, even when these names used non-Latin alphabets (such as العربية [Arabic]).

Machine translation (Level 0 support) is offered on the website, as required by the Multilingual Information Provisions Policy; however, even though Google Translate now offers translation between English and more than 100 other languages, only 51 languages are incorporated into the City of Toronto’s “Google Translate Tool.” Since many of these 51 languages are not among those most commonly spoken at home in Toronto, the city has likely not updated its machine translation page in some time, nor has it customized the languages to match the needs of its residents.[13] The city also clearly differentiates the translations produced by the Google Translate tool from the ones produced by the city, since the city’s Google Translate page includes a disclaimer stipulating that the city is not responsible for the accuracy of the translation (City of Toronto 2020). For this reason, when City of Toronto translations are discussed in the rest of this paper, I am referring only to the human-produced translations disseminated by the city, and not to the content that could be produced via the city’s integrated Google Translate tool.

In terms of translated documents, the City of Toronto website offered only PDFs, which we shall look at in more detail in the next section. The website also included three embedded YouTube videos with dubbing or ASL interpretation. Of these multimedia files, one, about newborn hearing screening, was dubbed (and captioned) in twelve languages. This video was the only instance of translation into an Indigenous language on the website: one of the dubbed languages was Ojicree, while the others were new immigrant languages (Arabic, Chinese, Farsi, Italian, Portuguese, Somali, Spanish, Tamil, Urdu, and Vietnamese) and French. The other two other videos, about voting in municipal elections, included ASL interpreting. These two videos were the only files found to offer support for ASL—though of course ASL support cannot be offered through the medium of PDFs, which were the only other type of translated documents on the website.

The only language to reach beyond Level 1 was French, since the website contained 18 pages either entirely in French or with both French and English content. These 18 pages discussed healthcare (information on long-term care facilities), childcare and parenting (information about childcare centres), education (information about health programs in schools), and employment (information about employment and social services locations).

2.4.3. Translated documents

Given that most of the translation on the City of Toronto website remained at Level 1, it is worth focusing on the types of documents that were available on the website. To facilitate the analysis, I have grouped the PDFs into ten categories, based on the content of the document. Each document was assigned to only one category. This did occasionally pose problems for documents with broad content, but I established criteria to systematically ensure that documents in each category would share the same characteristics and would not overlap with another category.

The translated PDFs fell into the following categories:

  1. Childcare and parenting documents were about infants and/or children and were aimed at parents or families, rather than educators (in which case they fell into category 8: education and teaching resources). Documents that discussed maternal health or healthcare for infants and children were classified as category 4 (healthcare) instead. Translated documents in the childcare and parenting category included a brochure on building resilient youth and a guide for parents who would be leaving their children home alone.

  2. Elections documents were related to municipal, provincial or federal elections, and included guidelines for voting in upcoming municipal elections.

  3. Immigration documents were aimed at people moving to the city for the first time. The one document that fell into this category was a newcomer welcome brochure.

  4. Healthcare documents discussed mental and physical health and wellbeing. Maternal healthcare—such as pamphlets about breastfeeding or emotional health during pregnancy—were common, but there were also documents about hygiene, such as a poster on handwashing.

  5. Emergency planning documents discussed fire safety or advised residents about how to prepare for natural disasters, medical emergencies and even (given Toronto’s proximity to the nuclear power plant in Pickering, Ontario) nuclear disasters.

  6. Urban development documents covered a range of topics, but were all related to city-led construction, transportation, and infrastructure projects. For instance, the city translated a poverty reduction strategy, construction notices, and a transportation master plan for a busy traffic corridor.

  7. Housing and shelter documents were of two kinds: those that targeted homeowners and renters, offering them information about city housing-related policies, and those that discussed emergency or homeless shelters. Three documents fell into this category: a guide advising homeowners about proper renovation processes, a form that renters could use to request services or repairs from their landlords, and a document explaining why a temporary shelter would be collecting personal information.

  8. Education and teaching resources were either documents targeted at educators (such as a guide to teaching elementary school students about puberty) or forms that could be filled out for various school programs (such as a grant application for a school food program)

  9. Legal documents were related to courts, legislation or bylaws. Only two documents fell into this category: a set of rules and procedures for tribunals and a form for requesting an extension on paying parking fines.

  10. Other was reserved for documents that fit into no other category. Only one document was included: a guide to using the Toronto archives to research a home’s history.

With the exception of French, the 28 languages in which the PDF documents were available are all new immigrant languages. Figure 1 compares the number of documents translated into each of the 28 languages with the 2016 census data on the number of people who speak these languages in Toronto. Note that in Figure 1, the left and right axes have different scales. The left axis depicts the number of translated documents on the website (grey bars), which ranges from 0 to 47, while the right axis shows the number of people in Toronto who reported speaking this language most often at home according to the 2016 census (black bars), which ranges from 2,995 (the number of Pashto speakers) to 171,165 (the number of Mandarin and Cantonese speakers).[14] The data is sorted based on the number of language speakers, but the X-axis remains constant, allowing us to visualize how the number of translated documents compares to the number of people who speak that language most often at home in Toronto.

Figure 1

Number of translated documents compared to number of people in Toronto who speak these languages most often at home[15]

Number of translated documents compared to number of people in Toronto who speak these languages most often at home15

-> See the list of figures

Figure 1 demonstrates that to some extent, as the number of people who speak a given language at home declines, so does the number of translations available in that language. There are several exceptions to this trend, the most obvious of which is French: almost all of the 52 documents were translated into French, even though French speakers make up a very small percentage of Toronto’s population: approximately 0.7% of Toronto residents reported speaking French most often at home in the 2016 census, about the same number of people who reported speaking Vietnamese, Gujarati or Arabic, and yet only 9 documents were translated into Vietnamese, 10 in Gujarati, and 17 into Arabic, compared to 47 in French.

The fact that more documents are available French than in other languages is not entirely unexpected, due to several factors. First, as discussed above, the City of Toronto’s current policy (as well as its previous policy) gives special status to French. For instance, critical information must be translated into the top ten languages spoken at home and French (City of Toronto 2017a: 3). Previously, the city recommended that translations into French be provided whenever public information on citywide issues were also translated into other languages (City of Toronto 2002: 2). This policy would undoubtedly have contributed to some of the translations into French, particularly for non-critical texts published prior to 2017, when the old policy was in effect.

Second, some of the 47 French documents were produced by the Ontario government, rather than the City of Toronto. Under the province’s French Language Services Act (Legislative Assembly of Ontario 1989/2019), the City of Toronto is a designated service area where residents are able to access services from the provincial government in both French and English, and documents from the province (such as the 2018 Candidates’ Guide for Ontario Municipal Council and School Board Elections, which is posted on the city’s Election Candidate Information page) would need to be produced in both official languages. Although the French Language Services Act does not oblige the City of Toronto to translate municipal material into French (Hébert 2016: 24), the city might still have to do so for reasons that are not part of either the Multilingual Information Provisions Policy or Bylaw 1176-2009. For instance, if Toronto offers a service or program on behalf of the province or has a funding agreement with the federal government, then the city would have to make material related to these programs and services available in both French and English (City of Toronto 2017c). This would seem to be the case, for instance, with a student nutrition program grant application form, since the grant was administered jointly by the City of Toronto and the Province of Ontario.

Finally, some of the French translations were disseminated but not produced by the City of Toronto, which means that the translation costs did not come from the city’s budget. For instance, the webpage where the French edition of a teacher’s guide on puberty and sexual health (City of Toronto 2017d) can be downloaded[16] has a statement thanking Public Health Ottawa for the French translation.

The differences in the number of translations for other languages is not as easily explained. At first glance, it is difficult to say, for instance, why the city has translated more documents into Arabic than into Gujarati or Vietnamese when there are almost the same number of speakers of all three languages. Likewise, almost the same number of people speak Tagalog, Tamil, and Spanish most often at home, and yet the number of translations differs markedly, with 17, 20 and 26 documents respectively available in these languages. Finally, languages such as Romanian, Turkish, and Armenian have more speakers than languages like Tibetan, Amharic, and Pashto, but no translated documents are available in the former while 1 to 3 documents are available in the latter. Taking a look at the breakdown of information categories for each of the languages, though, does help contextualize these figures.

Figure 2

Translated documents on the website, by category and language

Translated documents on the website, by category and language

Due to space constraints, language names have been replaced with their respective ISO 639 codes.[17] There were no separate codes for Farsi and Dari, so Dari has been left unchanged. AR (Arabic), ARC (Assyrian Neo-Aramaic), BN (Bengali), BO (Tibetan), Dari, EL (Greek), ES (Spanish), FA (Farsi), FR (French), GU (Gujarati), HI (Hindi), HU (Hungarian), IT (Italian), KO (Korean), PA (Punjabi), PL (Polish), PS (Pashto), PT (Portuguese), RU (Russian), SO (Somali), TA (Tamil), TL (Tagalog), UK (Ukrainian), UR (Urdu), VI (Vietnamese), ZH (Simplified & Traditional Chinese)

-> See the list of figures

As Figure 2 illustrates, the category with the most translated documents was healthcare: 14 of the 52 documents fell into this category; however, only 17 of the 28 languages had a translation of a health-care-related text, and these languages are typically the ones with the most translations. Election-related material, which comprised 10 of the 52 documents, and housing and shelter documents, which comprised only 3 of the 52 documents, were translated into the most languages. In fact, only four of the 28 target languages—Hindi, Hungarian, Dari, and Amharic—did not have a translation of at least one of the election documents, while only five target languages—Italian, Assyrian Neo-Aramaic, Pashto, Tibetan, and Ukrainian—did not have a translation of a housing or shelter document.

As noted earlier, the translation of information related to municipal elections falls under Bylaw 1176-2009 rather than the Multilingual Information Provisions Policy. Under the bylaw, the city may translate election information when at least 2% of a ward’s population speaks a given language, which means texts can be translated into languages that are not among the most widely spoken city-wide. This may therefore explain why translated documents are available in languages such as Assyrian Neo-Aramaic, Tibetan, and Pashto but not in more widely spoken languages such as Romanian, Turkish, Armenian or Albanian, since the only documents translated into Assyrian Neo-Aramaic, Tibetan, and Pashto were elections-related material such as what a voter card looked like and how to cast a vote in the 2018 municipal election.

It is also evident that some categories—namely education and legal forms—were translated only into French and no other languages, while only one category—urban development—was not translated into French (though it was translated into eight other languages). This again highlights the competing approaches in the city’s translation policy—namely a clash between the translation “as a right” and translation “as accommodation” (see Cordoba Serrano and Diaz Fouces 2018: 6-10). Because education in Canada is a provincial responsibility, and Toronto is a designated area where the province of Ontario must provide French-language services, many educational forms must be available in French, and thus translation is a right, regardless of the number of people who speak French in the city. By contrast, texts about urban development, a municipal responsibility, would not have to be translated into French, and given that the information in this category is not critical, the city’s translation policy treats French like any other language spoken in the city. In this case, translation is an accommodation: the number of speakers seemingly did not qualify French for a translation, since the urban development documents were translated only into six of the seven most frequently spoken languages.

3. Translation and linguistic justice on the website

This final section will consider questions related to linguistic and translation justice as they apply to the City of Toronto’s translation policy and the website.

In Section 2.3, I argued that the City of Toronto’s translation policies largely prioritize instrumental or non-identity interests (De Schutter 2017: 17) over identity interests, given their emphasis on communicative efficiency. This same focus on communication, rather than identity promotion or symbolic affirmation (Patten 2001) is also visible in the website. The translated webpages, multimedia, and PDFs typically offer important information about health, safety, and elections, rather than material that celebrates cultural heritage or helps a particular language to flourish in the city. Virtually no translations are related to recreational programs, the arts or cultural history.[18] Moreover, in the translation policy itself, languages (other than French, to a limited extent) are not prioritized based on their symbolic connection to Canadian culture or as a way to empower historically disadvantaged language groups. The languages into which translations are produced are those most frequently spoken at home, rather than those that are in danger of disappearing in the city—Indigenous languages, for instance.

Communicative efficiency is also visible in the city’s decision to embed translated documents on the website within an English interface and English navigation menus. While some pages were available in French, the navigation menu was always in English, as was the case for the PDF documents. In addition, the website lacks a page listing all translated documents, and the integrated machine translation options do not overlap with many of the city’s most commonly spoken languages. This means that residents are being only partially accommodated through translation and are expected to be able to function in English or to have someone help them navigate the website.

While the policy documents closely reflect the city’s instrumental interests, their implementation is not always consistent. The city attempts to strike a balance between reaching as many language groups as possible and minimizing financial costs by translating only certain types of texts and only if the number of speakers of a language reaches a certain threshold (for instance, a percentage of a ward’s population). However, some Torontonians (such as Arabic speakers) have access to more documents in the language they speak most often at home, even when the same number of Toronto residents speak one of the city’s other languages (like Gujarati and Vietnamese). The kinds of information that are translated for Toronto residents also varies by language, with only some language speakers having access to, say, elections materials and emergency planning documents.

As noted, the city of Toronto’s website generally did not reach beyond the first level of website localization, in the revised Jiménez-Crespo (2013) model. The 0-5 “levels of localization” scale implies that the ideal situation would be Level 5, namely a fully functional mirror site, but how would this be feasible, from a financial standpoint, for the 28 languages in which the city has made documents available, never mind the other languages spoken in the city? Obviously, in cases like Toronto, where the non-official language communities are varied and numerous, and when instrumental, rather than identity interests, are the focus, aiming for a high level of localization for non-official language groups is not communicatively efficient. Thus, Levels 0-3 might be appropriate for a municipality that aims to maximize the number of languages in which information can be made available. To help ensure that translated documents are more readily accessible, these minimal levels of localization could be combined with more effective navigation, such as a link in the navigation menu to a page with a list of all translated documents, along with a link to each document written in both English and the native name of the language. In addition, building on the recommendation in Carroll’s (2010: 388-389) study of Japanese municipal websites, documents could be made available in simplified English, which would help facilitate communication with all recent immigrants whose first language is not English.

Although instrumental interests are the focus of the city’s translation policies, identity interests are also represented to some extent, offering an example of the dual identity/instrumentalist position described by De Schutter (2017: 28-29). The city’s previous policy more explicitly emphasized both identity and instrumental interests, since French was offered certain protections that other languages were not due to its official status at the federal level (Hébert 2016; City of Toronto 2002). However, even the current policy gives special status to French in certain circumstances (critical information, and marketing material related to Canadian history and culture). As we have seen, this likely contributed to the number of translations into French being considerably higher than the translations into any other language, despite the small number of people who speak French at home in Toronto.

By contrast, under the existing translation policies, the City of Toronto’s translation efforts are unlikely to include Indigenous languages. After all, in the Canadian census, the number of First Nations people who report being able to speak or understand an Indigenous language has decreased from fewer than 30% in 2006 to only 20% in 2016 (Duff and Li 2009: 3; Anderson 2018; Assembly of First Nations 2019: 6[19]).Prioritizing languages that meet a certain threshold in the census data, as the City of Toronto (2017a) Multilingual Information Provisions Policy does, means that endangered languages are unlikely to be a focus of the city’s translation efforts. And yet, the city’s objective in the Multilingual Information Provisions Policy is to “reflect the linguistic diversity of the city” (City of Toronto 2017a: 1). This could be better achieved with a policy that includes ways for the city to provide information in official, new immigrant, and Indigenous languages. As the Assembly of First Nations (2019: 7) has argued, effective language policies are one of the elements that are crucial to the survival and revival of indigenous languages. Incorporating a provision for indigenous languages within its Multilingual Information Provisions Policy would be in the city’s best interest, because if language diversity is valued, “a policy that promises to deliver more diversity should be preferred over one that promises to deliver less” (Grin 2003b: 185), provided the financial and symbolic costs are roughly the same.

The city could aim to reflect a broader range of linguistic diversity while also minimizing the financial costs by sharing resources with other municipalities or even with other local or regional partners, such as universities and language revitalization groups, as suggested by the Assembly of First Nations (2019: 37). Evidence of collaboration is already visible on the website, since several of the French translations featured there were not prepared by the city but rather by the province or other agencies, such as Public Health Ottawa. A similar practice could be followed for Indigenous languages and less widely spoken new immigrant languages. For instance, two municipalities could write an emergency planning document and share the translation costs, or a municipality with a significant number of speakers of a particular language could share relevant documents with a municipality that does not have enough speakers of that language to justify the translation or writing costs.

Other ways of balancing competing instrumental and identity interests could be explored more fully in the future by studying municipal websites in areas where the linguistic, social and political contexts differ from those in Toronto: bilingual municipalities such as Moncton, New Brunswick; municipalities such as Richmond, British Columbia or Markham, Ontario, where most non-official language speakers speak the same language; small municipalities with fewer resources than the City of Toronto; municipalities in Quebec, where French (rather than English) is the official language, etc.