The focus of this book is to see how old and new technologies can be combined in a practical manner for endangered language maintenance and research. It is aimed at a wide readership including computational linguists, linguists, language documentation experts, language educators, digital curators, those who work with endangered languages, and the generally curious.
Jones has combined papers by researchers from Europe, Asia, Africa, and North and South America. Part 1 looks at creating new technologies for endangered languages, while part 2 looks at how new technologies are applied to endangered languages.
Jones outlines the precarious position of many of the world's languages. She reports that the availability of new technologies has the potential to improve many tasks in the field of endangered languages, including documentation, analysis, and revitalisation. She sounds a cautionary note, however, saying ‘easier’ does not automatically imply that things are better. If the native speaker becomes more invisible by the use of new technologies, then surely something is being lost in the process.
In the introduction, Nicholas Ostler provides an interesting overview of the world's languages and states that there are four areas where technology will have an impact on language: theoretical studies, support for recorded media, mechanisms for improved language access for speakers, and mechanisms for improved language access for nonspeakers. He speculates that perhaps in the future new technologies, particularly machine translation (MT), may facilitate interlinguistic communication, and thus make ‘smaller’ languages less threatened.
In part 1, Aimée Lahaussois describes the design and development of a prototype corpus of three endangered languages in eastern Nepal, namely Khaling, Thulung, and Koyi. The approach is to include material from a similar story in all three languages in a comparable corpus, and she outlines how she established a typology of similarities for her analysis. She outlines the tools that were developed to view the corpus (Integral Text View and Similarity View) and analyse it. Although the prototype corpus is small, it has yielded some interesting results.
Sjef Barbiers discusses the establishment of the European Dialect Syntax project (Edisyn), a documentation and research infrastructure for endangered dialects. He outlines the methodological requirements used for the collection of data and lists several criteria that should be met. He suggests that English glosses, parts-of-speech (POS) tagging, and geo-referencing should be used to enrich the data. Barbiers notes that is it important that projects such as Edisyn integrate with other similar projects such as LLMap (linguistlist.org), SSWL (Syntactic Structures of the World's Languages) and WALS (World Atlas of Linguistic Structures).
Hugh Patterson looks at the under-reported issue of keyboard design for endangered languages and focuses on four languages in the Me'phaa language family. He notes that the design must take into account technical differences between different operating systems, keyboard layout of the locally dominant language, the keystroke frequency, and Unicode compatibility issues. Patterson says that there are technological limitations, sociolinguistic pressures, and user experience issues to be dealt with in designing a keyboard for endangered languages.
Matt Coler & Petr Homola state that while MT can provide insights on language and facilitate communication, neither Rule Based Machine Translation (RBMT) nor statistical MT are suitable for endangered languages. They provide an introduction to using Lexical Functional Grammar (LFG) for Aymara. They state that the highly polysynthetic nature of Aymara would present challenges to traditional statistical MT tools. They offer their approach as an example of how it is possible to develop a tool for the documentation of endangered languages and related dialects.
Dorothee Beermann outlines the organisation of the linguistic work flow and suggests that internet-based tools can help linguists with data management, including accessibility and rights issues. She shows how Interlinear Glossed Text (IGT) can be used, giving examples for Akan and Kwa. She explains how a collaborative data management system (CDMS) and virtual research environments (VREs) can facilitate data management, and promote both the standardisation of linguistic annotations and open access. She says linguists have a responsibility to document vulnerable languages and make the data accessible.
Russell Hugo argues that it is important for endangered-language educators to have a long-term perspective, even when developing for the short-term. He lists several key questions that should be considered including some relating to goals, content, and pedagogy. Hugo provides an example of how to use a learning management system (LMS), in this case Moodle, for the Xinkan language family and for First Nation languages in Canada. He concludes by stating that the limited resources available to language revitalisers should be focused on tasks more important than developing a language-learning programme from scratch, especially when suitable resources may already exist.
In part 2, Bernard Bel & Médéric Gasquet-Cyrus outline digital curation and event-driven methods in the area of endangered languages, providing an example of digital curation of the patois spoken in Valjouffrey. They state that while recoding speech is important, methodological, technical, and ethical issues should also be considered. Their digital curation framework aims to preserve data, metadata, information regarding access rights, and data usability. They emphasise that ethical matters must be taken into consideration beyond what would legally be required and point out that a long-term vision must be embedded from the start of a project.
Anthony Scott Warren & Geraint Jennings provide an overview of Jèrriais and the difficulties small teams can have when trying to assess which (new) technologies may be useful for language revitalisation. The authors note that the development of a website for the language provided the starting point for a primitive corpus of the language. Twitter, Facebook, Wikipedia, and a virtual learning environment (VLE) have all provided new outlets for the language. Scott Warren & Jennings note that, as the internet has become more accessible to nontechnical users, social networking has become a way of interacting with others, and that they are trying to leverage that for Jèrriais.
Tjeerd de Graaf, Cor van der Meer, & Lysbeth Jongbloed-Faber discuss new technologies and Frisian. Popular resources including an online language-learning facility, education materials for different devices, a Twitter account, a Facebook page, and a Wikipedia presence. The Fryske Akademy (Frisian language academy) uses new technologies for various purposes—all with the aim of encouraging people to use the language in everyday communication. The authors explain how their work with the Mercator European Research Centre on Multilingualism and Language Learning provides an example for other minority languages in the European Union.
Cecilia Odé provides an overview of the native-speaker viewpoint and what native speakers require from language researchers. She notes that the language databases produced with speakers from the Tundra Yukaghir (TY) speech community are useful for linguists but they are of no use to the community itself. Even the provision of material on the internet for free is not of use, as most community members do not have access to the internet. Odé explains the importance of editing video recordings and providing metadata for materials and presenting data in a nonscientific manner. She summarises by saying that new technologies can raise awareness of endangered languages in the speaker community.
In the final chapter, Jeffrey E. Davis provides an overview of the documentation of American Indian Sign Language (AISL). He looks at how new technologies can be used for data collection and annotation of the language (with subtitles, voice-over, and slow motion). He discusses the challenges that arise when working in the field, the multilingual/multicultural/multimodal nature of the work, and the need for interdisciplinary collaboration between the various participants in the process. One of the main strengths of the project is the collaboration between the native AISL consultants and other researchers and how all parties learn how to work together in multidisciplinary teams.
This book makes for very interesting reading. Anyone who is interested in endangered languages, new technologies, or computational linguistics will find something of interest. Jones’ preface provides a succinct overview and sets the tone for the reader, while Ostler provides food for thought as to the future of the world's languages.
Part 1 of the book is more technical than part 2. Some of the chapters are more accessible to the nontechnical reader than others. Lahaussois’ chapter on the Kiranti comparable corpus provides an overview of the difficulties of working with languages with a mainly oral tradition and little in the way of language documentation. Barbiers works with a different type of endangered language and provides some suggestions for other endangered languages. Who would have thought of the importance of the keyboard layout for a language community (Paterson)? Coler & Homola see the potential benefit of using an LFG approach for low-density languages such as Aymara. Beermann reminds the reader of the importance of making the data convertible. Hugo provides a comprehensive overview of the questions to ask when designing a language-learning resource for an endangered language.
Part 2 is more comprehensible to the nontechnical reader. Bel & Gasquet-Cyrus highlight the importance of digital curation, not just documentation. Scott Warren & Jennings provide useful insights into an endangered language starting to make use of new technologies (Jèrriais) while de Graff, van der Meer, & Jongbloed-Faber show what is possible (Frisian). Odé makes some very pertinent points and reminds the reader that what is useful to the academic may have little or no value for the native speaker. Davis’ contribution on American Indian Sign Language covers an extensive amount of research and provides plenty of food for thought.
The strength of the book is the range of languages covered, the levels of endangerment of those languages, the new technologies used, and how they are used. The reader can learn lessons from some of the less endangered languages and those from better economically resourced parts of the world. The positive messages from some of the language communities are inspiring and the creativity and determination of the researchers in the face of many challenges is impressive. They are aware of the importance of their work, the responsibility it brings, and the privilege of being involved in the process. This book encapsulates the diversity of those working in the area, and there is something there for most readers.
