Introduction
Bhutan is a landlocked country located between India to the South and China to the North. Bhutan's population is 735,553 as of 30th May 2017 and a projected population of 770276 as of 2023 (National Statistics Bureau [NSB] 2023). Linguistically, Bhutan is diverse with 19 different languages spoken across the country (Tshering and van Driem Reference Tshering and van Driem2019). Dzongkha is the national language and is widely spoken by the majority of Bhutanese. However, English is used as a medium of instruction in Bhutanese schools. English and Dzongkha are used together as an official language of correspondence and administration (van Driem Reference Van Driem, Aris and Hutt1994). Hence, it could conceivably be put forward that the majority of Bhutanese are bilingual or multilingual in general. Moreover, it indicates the presence of multilingualism in Bhutan. However, a detailed study of the linguistic landscape (LL) of shop names is not available for Bhutan. The present study aims to fill this gap. So far, this paper has focused on the linguistic background of Bhutan. The following section discusses the LL scholarship and the purpose of carrying out the present study.
In recent years, LL has received increased attention across a number of disciplines. In addition, recent trends in multilingualism have led to a proliferation of studies that explore the LL of shop names (Nikolaou Reference Nikolaou2017), street signs (Amos Reference Amos2017) and signboards (Woo and Nora Riget, Reference Woo and Nora Riget2020). The central problem to be researched in the proposed study is the lack of study on the LL of shop names in Bhutan and the relationship between LL, language policies, and signage guidelines. As a basis for the study, I have identified the problem to be twofold. First, the prevailing issue identified is the lack of research on the LL of shop names in Bhutan. Previous studies have documented LL of shop names, street signs, and signboards in Asian countries including Thailand (Savski Reference Savski2021), Singapore (Shang and Guo Reference Shang and Guo2017), Korea (Tan and Tan Reference Tan and Tan2015), Malaysia (Coluzzi and Kitade Reference Coluzzi and Kitade2015; Woo and Nora Riget Reference Woo and Nora Riget2020), Japan (Backhaus Reference Backhaus2006b), and Indonesia (Sakhiyya and Martin–Anatias Reference Sakhiyya and Martin–Anatias2020).
Second, the prevailing issue identified is the lack of LL evidence and information on the relationship between the shop names, language policies, and signage guidelines. For example, the size and layout of shop names, and the code preference in terms of language choice. To address this issue, this study utilizes techniques advocated by Scollon and Scollon (Reference Scollon and Scollon2003) to identify code preferences in the shop signs. In addition, this study uses Signage Guideline 2017 (for more detail, see Thimphu Thromde [TT] 2017) and the National Policy and Strategy of Dzongkha Development (see Royal Government of Bhutan [RGoB] 2012 for more detail) to examine shop signs.
This paper aims to examine the LL of the shop names in Bhutan. More specifically, it evaluates LL, language policy, and signage guideline practices in the shop names. The key research questions of this study are (1) how are languages displayed on the shop names in Thimphu? (2) how are languages laid out on the shop names in Thimphu? and (3) how are shop signs laid out as per language policies and signage guidelines? To answer these questions, this study uses quantitative data. It is hoped that the findings of this study contribute to the field of bilingual and multilingualism in Bhutan. In addition, this study offers important insights into the LL of shop names and how they fit into signage guidelines and language policies.
This paper begins by providing the background on the type of languages spoken in Bhutan. It will then go on to review relevant studies on LL and language policy, and Bhutan's policy on shop signs. In the methodology section, I explain the data collection procedure and interpretation of the data. The next part summarises the main findings of this study and discusses the implications of the findings, and future research into the LL of the shop signs. The final section ties together the key findings, and provides recommendations for practice and policy, and further research work.
Literature review
Bhutan's policy on shop signs
The public signs and signboards in Bhutan are governed by Thimphu Thromde's [TT] (2017) guideline entitled ‘Signage Guideline 2017’. However, on the one hand, commercial shops established in recent years have adopted a bottom-up policy thus neglecting the signage guidelines. On the other hand, government-owned shop signs follow a top-down policy using signage guidelines. Therefore, there is a need for a clear policy with strict monitoring from government agencies. For example, the signage guideline clearly states that Dzongkha has to be a signage language along with English. Most importantly, the first letter on top of the signboard has to be Dzongkha. In addition, the guideline ensures that the Thimphu municipal office must maintain uniform signage with proper Dzongkha spelling. However, if there is any technical need in terms of Dzongkha spelling then Dzongkha Development Commission (DDC) must provide technical support. Similarly, according to the National Policy and Strategy of Dzongkha Development (RGoB 2012), public signs, such as shops and signboards must be written in Dzongkha and English. For example, DDC (2013) conducted a study to find Dzongkha usage on public signs and signboards. The study found that most public signboards have errors in Dzongkha spelling, punctuation, and direct transliteration. Based on the findings, DDC recommended that signboards should have: 1) Dzongkha text above English; 2) Dzongkha fonts bigger or equal to the English fonts; and 3) error-free Dzongkha letters on the signboards. The evidence reviewed here suggests a pertinent role for the municipal office and relevant agencies to monitor and strictly adhere to the signage guidelines and the language policy as mentioned above.
Linguistic landscape and language policy
In recent years, several studies (e.g., Han and Wu Reference Han and Wu2020; Manan et al. Reference Manan, David, Dumanig and Naqeebullah2015; Sakhiyya and Martin–Anatias Reference Sakhiyya and Martin–Anatias2020; Savski Reference Savski2021; Woo and Nora Riget Reference Woo and Nora Riget2020) have explored the relationships between LL and language policies when there is an explicit language policy on the shop signs. One of the most detailed studies of linguistic landscape and language policies comes from Hult (Reference Hult, Tollefson and Pérez–Milans2018). These studies have shown that most of the shop signs were non-compliant with language and signage policies imposed by the government. For example, Manan et al. (Reference Manan, David, Dumanig and Naqeebullah2015) examined the LL of Kuala Lumpur to establish a relationship between LL and language policies with a particular attention to top-down and bottom-up language policies. The study concluded that English remains a prominent language choice on the signboards despite the government's strict implementation of its signage policy. However, the study argued that defiance of the signage policy could be attributed to a multilingual society and a multi-ethnic population. Adopting a similar position, Savski (Reference Savski2021) carried out a study to understand the relationship between the language policy and the LL of Hat Yai in Thailand. Adopting a case study approach, the study collected shop signs and cafeteria menus from two sites, namely, the commercial district and the university cafeteria. The findings suggested that although there is an institutional mandate to follow policies on the semiotic structure, most of the semiotic structures were defying the signage policy, thus, indicating resistance towards the instructional signage policy. Moreover, the study argued that local agencies subvert their language policy when the semiotic structures are bottom-up. In a similar vein, Han and Wu (Reference Han and Wu2020) examined the relationship between language policies and actual practices in implementation on the signboards. The findings indicated that the officially LL creators have to strictly follow language policies. However, the study reported that LL creators use different tricks to avoid relevant language policies, especially signage guidelines by using a bottom-up policy in a few instances. This creates conflict and tension between language policies and LL. Similarly, results of Sakhiyya and Martin–Anatias’ (Reference Sakhiyya and Martin–Anatias2020) study indicated that government signboards follow top-down policies whereas private signboards were non-compliant with language policies and challenged the government signage policy. However, a notable example of an LL study where there is a correlation between signboards and language policies is Woo and Nora Riget's (Reference Woo and Nora Riget2020) study on signboards at Kuala Lumpur international airport in Malaysia. The study found the signboards at the international airports were displayed as per the government guidelines.
On the other hand, there are a few studies (e.g., Kasanga Reference Kasanga2012; Taylor–Leech Reference Taylor–Leech2012) that examined the relationship between language policy and LL where a connection is implicitly stated. Many of these studies have reported a connection between LL and de facto policies and recommend changes to language policies based on the findings. Kasanga (Reference Kasanga2012) demonstrated that in Cambodia the national language, Khmer, is dominant on signboards followed by English. However, the visibility of English is notable considering the prohibition of the use and teaching of the English language as per the language policy. Similarly, Taylor–Leech (Reference Taylor–Leech2012) analysed semiotics specifically public signboards in Dili, Timor-Leste. The most striking result to emerge from the data was that the national language was not visible on public signs. Therefore, the study recommended that greater efforts are needed to ensure higher visibility of the national language on public signs.
English language in Asian linguistic landscape
There are many well-documented studies on the dominant use of English in Asian LL, e.g., Malaysia (Manan et al. Reference Manan, David, Dumanig and Naqeebullah2015), Singapore (Tang Reference Tang2020; Zhang et al. Reference Zhang, Seilhamer and Cheung2021), Thailand (Thongtong Reference Thongtong2019), and South Korea (Lawrence Reference Lawrence2012). The studies presented thus far provide evidence that English has high visibility in shop signs, advertisements, street names, road signs, and public signs. One such study was conducted by Thongtong (Reference Thongtong2019) to examine the choice of languages used on the signboards in Thailand. The study found that English is a dominant language for monolingual, bilingual, and multilingual signs. In the same vein, Manan et al. (Reference Manan, David, Dumanig and Naqeebullah2015) explored the LL of the commercial district in Kuala Lumpur, Malaysia. The most obvious finding to emerge from the study was that English surpassed other languages; moreover, more space was allocated to the English language. Among the published work on the LL of Singapore, two studies drew our attention. The first one is Zhang et al.'s (Reference Zhang, Seilhamer and Cheung2021) study on LL of Chinatown in Singapore. Interestingly, the study was on Singapore's Chinatown nocturnal LL. In other words, data collection was carried out at night focusing on the signs glowing at night. Results revealed that English was widely used in most signs. The second study was Tang's (Reference Tang2020) recent investigation on language dominance in multilingual Singapore. The findings accord with the earlier study, which revealed that English was a dominant language and appeared 96% on the signboards. Adopting a similar position, Lawrence (Reference Lawrence2012) examined the usage of English in Korean LL. The findings indicated that English was the dominant language on public signs in South Korea. Overall, these studies suggest that English has a high visibility, and that it is the preferred language, and it is more dominant compared to other languages.
Methodology
Data collection
A quantitative approach was employed for the present study. Recent studies have used photographs of shop signs to analyse signage. Therefore, taking photographs of shop signs has become a popular data collection method for the LL studies (see Gorter and Pütz Reference Gorter, Pütz, Pütz and Mundt2018, for more detail). The photographs of shop signs were taken in October 2021. The photographs were taken daily for one month to capture all the possible shop names without omitting a single shop in the study area. While surveying the study area, I have noticed that several shops have more than one shop sign. In that case, I have taken a photograph of the shop names that were displayed on the shop front and a conspicuous one. However, when front and conspicuous signs were not found, the photo of inner and door signs was accounted for the shop signs. For the present study, I have collected 518 shop signs from the main street (Nordzin Lam) in the capital city of Bhutan. The data consists of photographs of signs and languages used.
Research site
Most studies on LL have focused on shopping streets as a survey area to collect picture materials for signage analysis (Backhaus Reference Backhaus2006a; Gorter and Pütz Reference Gorter, Pütz, Pütz and Mundt2018; Hult Reference Hult2014). For the present study, I have selected the main street of Thimphu city with shop signs in and around Nordzin Lam (see Figure 1). This area was chosen due to the high concentration of shops and commercial outlets. For many Bhutanese, Nordzin Lam is considered a shopping paradise due to its walking distance from most of the capital residents. Nordzin Lam stretches for 1.9 kilometres. Many shops offer both goods and services. The goods sold in shops are electronics, footwear, furniture, garments, vegetables, fruits, food, drinks, and so on. The services provided by the shops are laundry, banking, coaching, consulting, hairdressing, and others. Due to the high concentration of various shop signs, Nordzin Lam was an ideal place for the LL study. Other places in Thimphu could have been chosen but most of the shops in these areas were scattered and not viable for the LL study.
Map of the Nordzin Lam (from Google Maps).
Data analysis
To answer the research questions, the common practice of LL analysis (Backhaus Reference Backhaus2006a; Hult Reference Hult2010; Seals Reference Seals2017) was used where shop names collected were categorized into monolingual, bilingual, and multilingual shop signs. The common languages used in the shop names were Dzongkha and English. Moreover, the photographs of the shop names were meticulously analysed to see whether they were designed and written according to the government-provided signage guidelines. For instance, languages used on the signboard, font size of the letter, and layout of the text.
Results
The most important finding is that English is de rigueur on all the shop names irrespective of monolingual, bilingual, and multilingual shop signs. In addition, Dzongkha has a high presence in the bilingual shop names, but English is the preferred language of the shop signs. This could be due to the guidelines and regulations imposed by the government. In terms of code preference, shop owners prefer Dzongkha on top of the other languages and are placed in a vertical position. However, there is no uniform font size to indicate owners' code preferences. The other languages, such as Korean and Japanese, have a low presence of bilingual and multilingual signs. Detailed explanation of the results will be provided in the following sections.
Overview of Languages Used on Shop Names
The languages of the shop names are categorized into monolingual, bilingual, and multilingual signs. For instance, multilingual shop name signs are those signs which have three or more languages used. As can be seen from Table 1, the bilingual signs appear 362 times and the monolingual signs 153 times. This data indicates that most of the shop names are bilingual and monolingual. These results, therefore, indicate that most of the shop owners follow the signage policy of using Dzongkha and English on their shop signs. On the other hand, only three shop signs appear multilingual. This indicates that multilingual signs are least preferred by the shop owners.
Overview of languages used in the shop signs
Monolingual shop names
In terms of monolingual shop signs, the data indicate that English is the common monolingual language used on the shop name signs. The number and percentage of monolingual shop signs are presented in Table 2.
Monolingual Shop Signs
What stands out in the table is that most shops prefer to use English as a shop signs language as compared to Dzongkha (see Figure 3). For instance, 98.7% of the monolingual signs are written in English. A possible explanation for this might be that majority of the Bhutanese are literate in English and can read English easily (see NSB 2022 and Ura et al. Reference Ura, Alkire, Wangdi, Phuntsho and Zangmo2023 for more detail). In addition, it might be due to the owners’ choices to make their shops easily known to the customers.
In contrast, the shop name sign that appears in Dzongkha is very low as compared to English. For instance, there are only two shop names written in Dzongkha (see Figure 2), consisting of 1.3% of the monolingual signs. However, there are still some English words used in the shop names, as shown in Figure 2. A possible explanation of our findings is that reading Dzongkha is challenging, and another possible alternative explanation of our findings is that there is limited Dzongkha terminology as compared to English.
Monolingual Dzongkha shop sign.
Monolingual English shop sign.
Bilingual shop names
For bilingual shop name signs, three bilingual shop signs are English–Dzongkha, English–Korean, and English–Japanese. In terms of shop names, most of the shops used bilingual English–Dzongkha combinations (see Figure 4). For instance, 360 shop names make up 99.4% of the bilingual shop signs (see Table 3). This indicates that bilingual English–Dzongkha is the dominant language for bilingual shop signs. In addition, it indicates that shop signs are aligned with the signage guidelines where Dzongkha has to be the signage language along with English.
Bilingual English–Dzongkha shop sign.
Bilingual shop signs
On the other hand, a single shop name sign was written in an English–Korean combination (see Figure 6) and an English–Japanese combination (see Figure 5) was also found in the Nordzin Lam. These findings indicate that bilingual English–Dzongkha is the preferred language choice. This may be due to the government policy on the mandatory use of English–Dzongkha combination shop name signs.
Bilingual English–Japanese shop sign.
Bilingual English–Korean shop sign.
Multilingual shop names
Multilingual shop signs constitute 0.6% of the overall LL of the Nordzin Lam. The multilingual shop signs are presented in English–Dzongkha-Japanese and English–Korean-Japanese combinations. As shown in Table 4, English–Dzongkha-Japanese shop signs appear two times and English–Korean-Japanese shop signs appear in only one shop.
Multilingual shop signs
Figure 7 presents the multilingual shop signs of English–Dzongkha–Japanese. For instance, Dzongkha is displayed on the top and English at the bottom of the signboard. Likewise, English and Japanese are used in the middle of the signboard to display the products sold by the shop. The multilingual sign in the picture also shows the shop owner's business strategy to attract customers because the Japanese ‘Uniqlo’ is the top-selling brand in Bhutan.
Trilingual Shop Sign of Dzongkha-English–Japanese.
Code preference of shop signs
For the present study, I am using a code preference system advocated by Scollon and Scollon (Reference Scollon and Scollon2003). The code preference takes place when there are multiple codes present in the same space; thus, creating a system of choice to make signs more appealing in terms of visuals and language. In this section, I looked at the size and layout of bilingual and multilingual shop names. For example, in Figure 4, the Dzongkhag font is written on top of the English to show the priority and prominence of the Dzongkha script over the other languages. The position of the script can also indicate the importance of the Dzongkha over other languages. Therefore, this section presents the comparison of font sizes of different shop names in bilingual and multilingual shop names.
Figure 8 shows that Dzongkha is presented in larger font indicating the prominence of Dzongkha in the bilingual and multilingual shop names. However, as per the data collected, there is an uneven font size in the Dzongkha script making it difficult to comprehend shop owners’ preferences in the language use. On the other hand, the smaller font size presented in English and other languages indicates supplementary information.
Larger Dzongkha font size over English.
In addition to font size, the layout of the shop names in the bilingual and multilingual words also shows the shop owners' code preference as well as preference in their language choice. According to Scollon and Scollon (Reference Scollon and Scollon2003), the preferred code is placed on the top when placed vertically and on the left when placed horizontally. To investigate the preferred layout of the shop names, I looked at the bilingual Dzongkha–English shop names.
For example, the shops owned by the government use their shop names where Dzongkha is placed on the top and English is placed below; the layout and font are also displayed as per the signage guidelines. However, the shops owned by private individuals prefer to place the Dzongkhag on top of the English and other languages but there is no uniform layout and font size thus neglecting the government guidelines and the language policy. The choice of larger English font shows the shop owners’ preferred codes. Taken together, these results suggest that preferred codes tend to appear in a vertical position with Dzongkha on top of other languages. A similar code preference was found in Hong Kong where English is placed on top or left with Chinese below or right and it is regulated tightly by the government guidelines (Scollon and Scollon Reference Scollon and Scollon2003).
Discussion
This study set out with the aim to examine the LL of Nordzin Lam in Thimphu particularly paying a closer attention to languages used on the shop name signs. The most important finding is that English is de rigueur on all the shop names irrespective of monolingual, bilingual, or multilingual shop signs. This result is contrary to the signage guidelines (TT 2017) and the national policy and strategy of Dzongkha development (RGoB 2012), which state that shop signs should be written in Dzongkha and English. In addition, Dzongkha has a high presence in the bilingual signs, but English is the preferred language of the shop names. In terms of code preference, shop owners prefer Dzongkha on top of the other languages and are placed in a vertical position. These findings are in accord with the signage guideline, in which Dzongkha is the mandatory language on the signboards where Dzongkha script should be bigger or equal to English script and should be placed above English.
The results of this study showed that 99.4% of the shop signs were bilingual indicating the presence of language vitality of different languages in the LL of the shop names. As stated earlier, the two official languages in Bhutan are English and Dzongkha. Similarly, the present findings revealed that English and Dzongkha are the two dominant languages found on most shop signs. Other languages, such as Korean and Japanese are also visible but rarely used as compared to other vibrant languages like English and Dzongkha. This also accords with the findings of Shang and Guo (Reference Shang and Guo2017) and Woo and Nora Riget (Reference Woo and Nora Riget2020), who found English as a vibrant language in most shop names and signs in Malaysia and Singapore. Moreover, the presence of English in monolingual, bilingual, and multilingual shop signs suggests that English is a lingua franca (Phuntsho Reference Phuntsho2013) in Bhutan. Another important finding was that most of the Bhutanese are English literate with higher proficiency in speaking and reading. This can be a possible reason why most of the shop signs are written in English to provide the necessary information and make it more accessible to target customers. These results reflect those of previous studies (Lawrence Reference Lawrence2012; Manan et al. Reference Manan, David, Dumanig and Naqeebullah2015; Tang Reference Tang2020; Thongtong Reference Thongtong2019; Zhang et al. Reference Zhang, Seilhamer and Cheung2021), which found that English has high visibility in shop signs, advertisements, street names, road signs, and public signs. Moreover, they found that English occupies a prominent position in any given LL. These factors may explain why English is positioned as a vibrant language in the Bhutanese shop names.
The most salient finding from the investigation is that bilingual shop names are the most preferred code by the shop owners. In particular, a bilingual English–Dzongkha combination was preferred instead of another language combination. As seen in other studies (Backhaus Reference Backhaus2006b; Sakhiyya and Martin–Anatias Reference Sakhiyya and Martin–Anatias2020; Tan and Tan Reference Tan and Tan2015), English finds a prominent place in any bilingual signs in Asian LL. For example, Backhaus (Reference Backhaus2006b) examined the multilingual signs in Tokyo. The study surveyed 28 areas and collected 2321 multilingual signs. Results revealed that English accounted for 97.6% of the multilingual signs. Similar findings were also reported in Sakhiyya and Martin–Anatias’ (Reference Sakhiyya and Martin–Anatias2020) study on the LL of three cities in Indonesia. LL corpus gathered from three cities showed that English remained the predominant language in bilingual signs. This preferred code is chosen to sell products. Moreover, English is used as a sign maker for wider communication to inform the customer about the product.
The commercial, religious area, office, school, and private property names in Bhutan are guided by the unique language and signage policy. There is also some similarity with shop names and signs. The findings of this study revealed that most of the shop signs were written in English–Dzongkha combination following a top-down policy. However, these findings are noticeable in the shops owned by the government and the old-established shops. The signs of recently opened stores and shops tend to follow a bottom-up policy, neglecting signage guidelines and the language policy. Most importantly, the Thimphu municipal office should maintain uniform shop signs as mandated by the guidelines, but it is a rather laissez-faire approach. This finding echoes previous studies (Han and Wu Reference Han and Wu2020; Kasanga Reference Kasanga2012; Manan et al. Reference Manan, David, Dumanig and Naqeebullah2015; Sakhiyya and Martin–Anatias Reference Sakhiyya and Martin–Anatias2020; Savski Reference Savski2021; Taylor–Leech Reference Taylor–Leech2012; Woo and Nora Riget Reference Woo and Nora Riget2020) highlighting the importance of how language policies shape the overall LL of any given area. This finding has a close correlation to that of Taylor–Leech's (Reference Taylor–Leech2012) study on the LL of the island nation of Timor–Leste. The study found that multilingual signs with Tetum, Portuguese, Indonesian, and English were prominent in the shop signs. However, the national language was invisible in shop signs. In the same vein, Woo and Nora Riget (Reference Woo and Nora Riget2020) conducted an LL study to examine the languages used in two airport signs. The study revealed that signs used in Malaysian airports are monitored and regulated by language and signage policies. Taken together, the LL of the present study indicates that there is a divergence between language use and standard practices. For example, new commercial shops and outlets use monolingual English only, which goes against the language policy and signage guidelines. Moreover, this study found that English has high visibility in monolingual and bilingual shop signs. For example, in bilingual shop signs, English characters outnumbered those of Dzongkha. In addition, there is growing concern over new shop signs deviating from the language policies and signage guidelines whereby shop owners create their shop signs in response to linguistic diversity.
Conclusion
The present research aimed to examine the LL of the shop names in Thimphu with a particular focus on Nordzin Lam. The results revealed that English is the dominant language in monolingual, bilingual, and multilingual shop signs. In terms of code preference, most of the shop signs placed Dzongkha on top of English. This finding is in accord with the signage guideline, in which Dzongkha is a mandatory language on the signboards. The Dzongkha script should be bigger or equal to the English script and it should be placed above English. However, size and layout were found to be inconsistent in most shop signs. In addition, the shops owned by the state follow a top-down policy whereas the shops owned by the private individuals follow a bottom-up policy indicating ineffective policy implementation by the municipal office.
These findings are relevant to both practitioners and policymakers to further debate and strengthen the signage policy. However, it is important to note that this study only aims to provide an overview of LL of shop names in Thimphu and excluded important details on shop names, e.g., language errors, etc. The findings in this study are subject to at least two limitations. First, with a small sample size, caution must be applied, as the findings might not be transferable to the whole country. Further studies need to be carried out to collect data from other areas in Bhutan. Second, the discussion of the findings was more speculative which might be biased. Further research should be undertaken to explore the attitude of the shop owners using interviews or questionnaires.
Notwithstanding these limitations, the present study offers valuable insights into the LL of Bhutan and this information can be used to develop targeted interventions aimed at promoting and strengthening the signage guidelines and the language policy in particular.
SINGAY is a PhD student at the Faculty of Wellbeing, Education and Language Studies, The Open University, Milton Keynes, UK. Prior to undertaking doctoral study, he worked as a researcher at the Centre for Bhutan and GNH Studies, Thimphu, Bhutan. His research interests focus on Translanguaging, EMI, Bilingualism and Language Policy. Email: singay.singay@open.ac.uk
Open access
