Promoting minority language use on bilingual Web sites
Daniel Cunliffe Hypermedia Research Unit, School of Computing, University of Glamorgan, Pontypridd, Wales, UK.
Abstract
The World Wide Web offers minority languages the opportunity to reach a wide audience at relatively low cost compared to traditional media. It can be argued that the presence of a minority language in this new mass medium will become as important to language survival as having a presence in traditional mass media. This paper focuses on bilingual Web sites, defined as sites which include information in two languages and whose readers include bilingual speakers. Whilst the context of the work presented here has been Wales and Welsh/English bilingualism, it has relevance for other language communities with similar characteristics.
The poor design of many current bilingual sites may have implications beyond restricting their immediate use by minority language speakers; it may reduce their willingness to use the language on other sites. It may also be perceived as reflecting the lack of value that particular organisations place on a language. To facilitate and promote the use of minority languages on a Web site it is necessary to consider not only the readers’ goals, but also the influence language will have on the way they achieve them.
Bilingual readers generally make two types of language related decisions when visiting a Web site; the initial selection of a language and whether to move from one language to another. In order to promote minority language use, rather than simply facilitate it, it is necessary to consider not only how readers can be encouraged to select the minority language, but also how they can be encouraged to continue using it during their visit. This paper discusses two possibilities that are currently being explored in a prototype bilingual Web portal; the inclusion of a subject specific bilingual dictionary and the ability to view a translation of a passage of text rather than having to switch languages completely. It is hoped that the provision of a more supportive language environment will encourage the use of minority language both when using the site and in general within the subject domain.
Introduction
One of Crystal’s six postulates for a theory of language revitalization is that “An endangered language will progress if its speakers can make use of electronic technology” (Crystal, 2000, p141). The World Wide Web is currently perhaps the most significant of these electronic technologies and it can be argued that it constitutes a new form of mass media. Where minority language speakers have affordable access to the Web, it offers that language an important opportunity. This is due to a variety of factors, including:
• • •
In many countries there is little state regulation of the Web. It is possible to reach a wide audience at relatively low cost compared with traditional media. The Web is increasingly used as an information source by children in education and in the home.
It can be argued that the presence of a minority language in this new mass medium will become as important to language survival as having a presence in traditional mass media.
There are already numerous examples of Web sites that act as single language ‘channels’ for minority languages (for example, Guardado, 1998; Falkena, 2001). Whilst this represents one model of the way in which minority languages might use the Web, this paper focuses on bilingual Web sites. For the purposes of this paper these are defined as sites which include information in two languages and whose readers include bilingual speakers. Whilst the context of this work to date has been Wales and Welsh/English bilingualism, it has relevance for other language communities with similar characteristics.
1
A bilingual Web site can:
•    Acknowledge the existence of a minority language by making it visible to both speakers and non-speakers. A site of this type may provide little or no actual content in the minority language, but may for example include a minority language welcome message on the homepage.
•    Raise awareness of a minority language by presenting information about that language to both speakers and non-speakers. A site which is about a minority language may provide little actual content in the minority language if it is aimed at majority language speakers.
•    Facilitate the use of a minority language by making information available in that language. A site of this type will passively support use by minority language speakers by providing quality content in the minority language and by adopting a design which reflects likely patterns of use by bilingual speakers.
•    Promote the use of a minority language by providing both information and language support. In addition to facilitating use of a minority language, a site of this type will include content and adopt a design which actively promotes and encourages the use of the minority language.
An important distinction between sites which acknowledge or raise awareness and those that facilitate or promote is that the latter support the minority language as a working language rather than a ‘heritage’ language.
The use of the Web has added impetus within Wales due to the UK Government’s commitment to providing electronic access to government services (including Welsh language provision), Welsh Assembly Government promotion of the Web through initiatives such as Cymru Ar-lein (a strategic framework for Wales in the information age) and the Welsh Language Act (1993) which requires the public sector to treat the Welsh and English languages on an equal basis when providing services to the public in Wales (including electronic services).
Williams (1993, cited in Aitcheson and Carter, 1999, p181) identifies five foci of social pressures for language change:
1.    Idealism – the vision of a fully rehabilitated threatened language, usually related to the notion of making language and nation coterminous.
2.    Protest – the mobilisation of agitation behind a lesser-used language. 3.    Legitimacy – securing normalcy in exercising language rights in selected domains. 4.    Institutionalisation – ensuring the language is represented in key strategic agencies of the State, for
example in law, education and public administration. 5.    Normalisation – extending the use of the language into an optimum range of domains, especially
the public sector.
Given the impetus for using Welsh on the Web, it would seem reasonable to expect the Web to be making a significant contribution towards enhancing the legitimacy, institutionalisation and normalisation of the Welsh language. However, studies of Web sites in Wales has shown that there is great variability in the usability, scope and quality of Welsh language provision, even on the sites of organisations with a statutory obligation under the Welsh Language Act (Cunliffe, 2001; Linguacambria Cyf., 2001).
Poor minority language support may be as a result of a number of different factors, including:
•    Lack of additional resources required to develop a bilingual site. •    Lack of an easily measurable return on investment which places a value on language. •    Recognition that majority language provision will meet the needs of the majority of users. •    Lack of value placed on minority language by organisations. •    Lack of understanding of bilingual issues by Web developers.
2
The lack of understanding of bilingual issues by Web developers can be broken down into a number of areas:
•    Uncertainty over whether the activities undertaken during the development process should be bilingual.
•    Uncertainty over whether creation in one language and translation into the other is an appropriate model for content development.
•    Failure to adopt appropriate overall structural solutions. •    Inappropriate design solutions that are not sensitive to language issues. •    Failure to consider bilingual issues during testing and evaluation. •    Failure to monitor long term use of a site with regard to language use.
The poor bilingual usability, limited scope of minority language material and the poor quality of that material on Web sites may have a number of implications:
•    Readers will be unable to use their minority language on the site. •    Readers may be less likely to use minority language provision on other sites. •    Readers may perceive that particular organisations do not value the minority language. •    Readers’ perceived status of the minority language may be lowered.
It is therefore possible that poor design and implementation of minority language provision may actually work against the use of and survival of that language, particularly where this is seen on sites representing important and influential organisations
Developing for bilingual users
In contrast to multi-language sites where the readers can be presumed to be monolingual, readers from bilingual communities will generally have some level of competence in their second language (and often a high level of competence). This makes designing Web sites for bilingual communities subtly different to designing multi-language sites. To facilitate and promote the use of minority languages on a Web site it is necessary to consider not only the readers’ goals (e.g. locating a piece of information or purchasing a product), but also the influence language will have on the way they achieve them. Important reader choice points are the initial selection of a language and the decision to move from one language to another (often from the minority language to the majority language). There is a need to understand how design and other factors (such as quality or topic) will influence the readers’ choices. In order to promote minority language use, rather than simply facilitate it, it is necessary to consider not only how readers can be encouraged to select the minority language, but also how they can be encouraged to continue using it during their visit. By providing a more supportive language environment it may be possible to encourage the use of minority language both when using the site and in other contexts as well.
Whilst there are numerous publications addressing the issue of designing Web usability and some on designing sites with content in more than one language, there is little guidance for the design of sites for bilingual users (though see IBIS, 2000; TBCS, 2000; Cunliffe, 2001; Linguacambria Cyf., 2001).
It is also difficult to generalise from existing practice, as published case-studies of bilingual developments are few in number and often do not discuss the bilingual issues in detail. Papers describing Web sites for bilingual users include Voge (1998) and Hendler (2000) who describe a Spanish/English health information site; Guanipa (2001) who describes a Spanish/English information and referral site for psychoeducational services; and Cunliffe et al. (2002) who describe part of the development process for a Welsh/English resource portal for Welsh medium Speech and Language Therapists.
Almost all the work in this area is based on anecdotal or experiential evidence and existing practice in Web design. This may be in part due to the inaccessibility of academic literature in the study of bilingualism, but may also reflect the limited amount of research into bilingualism and new media.
3
Case study
The site, currently under development for the Resources Group of the Welsh Medium Speech and Language Therapy Committee, is a resource portal for Speech and Language Therapists working with Welsh speaking clients. It is intended to form part of the solution to perceived difficulties in sharing locally developed resources at a national level. Welsh language and bilingual Welsh/English therapy information and clinical resources will be listed in the portal and, in the majority of cases, will be available to download.
It is hoped that the resource portal will facilitate Welsh medium therapy by making existing resources more accessible and by sharing research and information. Reducing costs and the duplication of effort should also enable more Welsh language and bilingual resources to be developed. Thus the promotion of the Welsh language can be seen as one of the principal objectives of the portal.
Readers making use of the portal can be expected to be mainly first language Welsh speakers (with English as a second language to a high degree of fluency), but second language Welsh speakers and some Welsh learners may also make use of it. During the course of the development it became apparent that commonly accepted English language terms for Speech and Language Therapy concepts often have no agreed Welsh language equivalent. The translation of complex concepts, such as ‘adult acquired disorder’ is difficult, requiring both specialist knowledge of the subject domain and specialist knowledge of the Welsh language. It was also noted that even where Welsh language terms do exist they were often not in common use if the working language of a department was English, or even if the working language was Welsh. Among first language Welsh speakers, specialist vocabulary may be limited and English terms are often simply ‘borrowed’ when required.
Given the limitations of the Welsh vocabulary within the subject domain, the high levels of English fluency within the anticipated readers and the absence of any statutory obligation, it may be tempting to suggest that a site constructed solely in English would be perfectly usable, cheaper to develop and easier to maintain. This is entirely true, but fails to consider the broader social and (quasi-)political context of Wales generally and the Speech and Language Therapy community in Wales in particular, who are keen to promote Welsh medium therapy.
Taking into account this wider perspective it was agreed that a fully bilingual site was the appropriate solution. It was also decided to conduct the development process itself bilingually in as far as was practicable (Cunliffe et al., 2002). In order to actively promote (and normalise) the use of Welsh in this domain, the site would feature a subject specific bilingual dictionary which would be developed as an authoritative source of Welsh language Speech and Language Therapy terminology.
The overall structural solution (in terms of pages and links) of the original prototype site is shown in figure 1. A hierarchy of topic specific monolingual submenus provide access to a range of monolingual and bilingual resources. Each monolingual menu is linked directly to its equivalent menu in the other language, allowing easy traversal between Welsh and English. The initial entry point to the site is a bilingual homepage containing links to the monolingual sub-menus in their corresponding language. This allows readers to select a language almost subconsciously and also reinforces the fact that the site is wholly bilingual. Each menu and sub-menu also features a direct link back to the homepage in both languages. The dictionary was available only from the homepage and featured both Welsh-to-English and English-to- Welsh. The emphasis in this prototype was very much on allowing the reader to switch easily between languages whenever they wished.
4
Bilingual homepage
Bilingual glossary
Monolingual submenus
Monolingual and bilingual resources
Figure 1: Structural solution for the first prototype.
Discussions with Speech and Language Therapists following the presentation of the prototype were very positive but raised the issue of the ordering of resource descriptions on the sub-menus. As the descriptions in each language version were ordered alphabetically, there was clearly no correspondence between the ordering of descriptions in each language. This would make it difficult, for instance, to change quickly to the English version to check the meaning of an unfamiliar Welsh technical term and then return to the Welsh version. It was felt that this might result in readers initially picking English, or in remaining with English following a language change. Clearly the environment was not as supportive as it had been envisaged, and whilst this first prototype could be seen as facilitating minority language use, there was little in the design to actively promote it.
The prototype has since been modified in order to provide a more supportive language environment. The new prototype preserves the abstract structural solution of the original, but has new options which it is hoped will promote the use of the minority language (Figure 2). Firstly the bilingual dictionary can now be accessed directly from every page in the site and it opens up in a new browser window, rather than replacing the current page (Figure 2, A). Thus the reader is able to preserve the context of their work whilst they check the meaning of an unfamiliar technical term. Secondly the reader is able to open up a new browser window from an individual resource description which contains the equivalent of that resource descriptor in the other language (Figure 2, C). This may be more useful in some circumstances, than checking individual words in the glossary. Again the use of a new browser window preserves the reader’s original context of work and allows them to directly compare the two language versions. The reader can also simply change languages (Figure 2, B) as they could in the original prototype, in which case the current page is replaced by its equivalent in the other language.
One issue that hasn’t been addressed in the new prototype is that of initial language selection. Whilst it would be possible coerce the reader into using Welsh initially by providing a Welsh only homepage with a link to an English homepage, this goes against the bilingual nature of the site and against the principle of supporting reader control. Instead the bilingual homepage has been retained, this gives slight precedence to Welsh in terms of its overall layout, but offers the reader a choice of language.
5
A
1
2
3
B
2
3
1
C
Resource descriptions in language L1, ordered alphabetically by L1
Topic specific bilingual dictionary
Resource descriptions in language L2, ordered alphabetically by L2
2
Individual resource description in language L2
Figure 2 : From the current page in language L1, the reader can: A open the bilingual dictionary in new browser window; B change the current page for the equivalent in language L2; C open up a new browser window containing the complete L2 equivalent of an individual resource description.
Particular attention will be applied to the fine detail of the site, for instance file names and directory paths will also be language specific, enabling readers to remember and correctly enter URLs and to edit the URL as a means of navigation. Page titles will be in their appropriate language to aid recognition in bookmarks / favourites lists. Overall by providing a more supportive language environment it is hoped to encourage the use of minority language both when reading the site and in general within the Speech and Language Therapy domain.
Future work
Although the site has been redesigned to promote minority language use, the success or otherwise remains to be seen. In order to be able to make an informed judgement, Speech and Language Therapists’ use of the site will be evaluated once it is launched to see if we can we verify our hypotheses regarding language choice and change. This will involve a combination of direct observation, screen capture and post event analysis with Therapists. Access logs for the site will be collected and closely analysed to see if these can provide any insight into unobserved reader behaviour in the natural context of use.
Currently the site has been implemented using only very simple technology, HTML and Cascading Style Sheets. In the longer term the site will be converted into a database driven application. This should allow easier maintenance and greater flexibility, plus the addition of a search facility. It also raises the possibility of including more sophisticated language support such as pop-up word level translations. The benefits of adding community facilities to the site, such as bulletin boards or chat rooms, will also be investigated. Again the issue of promoting minority language use within these communal facilities will be considered.
6
The development of the site has raised other issues which can only be investigated in a broader context. In particular whether there are benefits in conducting the development of a bilingual site bilingually. There are significant overheads in conducting development activities bilingually in terms of preparing resources, translating, conducting activities and analysing the results, particularly if the Web development team is not itself bilingual. This is related to the question of whether developing content in one language and translating to the other is an appropriate model for bilingual sites.
This paper has raised the possibility of developing Web sites for bilingual readers which go beyond simply facilitating minority language use and which make some steps towards actively promoting its use through the provision of a more supportive language environment. Further work on the theoretical underpinnings of bilingualism in electronic environments and practical work evaluating the perception and use of Web sites for bilingual readers is required before the full benefits of the Web for minority languages can be realised.
References
Aitcheson J. and Carter H. (1999). “Cultural empowerment and language shift in Wales.” Tijdschrift voor Economische en Sociale Geografie, 90 (2): 168-183.
Crystal D. (2000). Language Death. Cambridge University Press.
Cunliffe D. (2001). “Bilingual usability of Unitary Authority Web sites in Wales.” Proceedings of The Web in Public Administration, EuroWeb 2001 Conference, O. Signore and B. Hopgood (Eds.), Pisa, Italy, 229- 244.
Cunliffe D., Jones H., Jarvis M., Egan K., Huws R. and Munro S. (2002). “Information architecture for bilingual Web sites.” Journal of the American Society for Information Science and Technology, 53 (10): 866-873.
Falkena O (2001). “www.omropfryslan.nl the development of a Frisian-language multimedia website.” Mercator Media Forum, 5: 42-47.
Guanipa C. (2001). “Amigos: A bilingual Internet-based resource for school counsellors.” Professional School Counseling, 4 (5): 362-365.
Guardado D. (1998). “Asturia – A head-start for Asturian on the Internet.” Mercator Media Forum, 3: 52- 59.
Hendler G. (2000). “NOAH – New York Online Access to Health: Connecting the community to reliable health information.” Health Care on the Internet, 4 (1): 3-16.
IBIS (2000). Interfaces for Bilingual Information Systems project reports. Available on-line: http://weblife.bangor.ac.uk/ibis/
Linguacambria Cyf. (2001) “Snapshot survey: Websites of organisations complying with statutory language schemes.” Report prepared for the Welsh Language Board. Available on-line: http://www.welsh- language-board.org.uk/uploads/publications/118.PDF
TBCS (2002). “Common look and feel for the Internet.” Treasury Board of Canada Secretariat website. Available on-line: http://www.cio-dpi.gc.ca/clf-upe/
Voge S. (1998). “NOAH – New York Online Access to Health: Library collaboration on bilingual consumer health information on the Internet.” Bulletin of the Medical Library Association, 86 (3): 326-334.
7