We thank Lenore Grenoble for her comments on our article. We are happy for this opportunity to clarify matters regarding ELCat's mechanisms, as well as the team's position on some crucial issues, chief among them being enumeration and commodification.
1. Grenoble cites Ashkun as she raises the question of enumeration, wondering how ELCat arrives at the number of 7,000 speakers, and what the various numbers mean. She also believes that ELCat appears to provide comparisons of sources but not new data.
ELCat aims to be an online resource for materials and research on endangered languages. It attempts to be an up-to-date listing of all relevant sources for each endangered language. The < 7,000 number actually comes from van Driem (Reference van Driem and Moseley2007). There are different numbers on each language page, because different sources may provide different speaker numbers. Users can scroll through to look at the different information provided by differing sources, noting the year of publication. Visitors to the site are also encouraged to participate. They can post language materials and suggest changes to existing information (subject to moderation). Users can gauge how much or how little information is available on particular languages, and as a result be encouraged to work on or help support those languages.
2. In response to Grenoble's comments, we clarify here our use of the term speaker.
In the current version of ELCat, which has been revised and improved since our original article was submitted, the LEI takes into consideration only native speakers when calculating the level of endangerment, but gives numbers of semi-speakers, second-language speakers, and rememberers to provide context and encourage active language revitalization efforts. For example, the language page of Northern Straits Salish reflects the number of native speakers (this figure contributes to the calculation of the language's level of endangerment), as well as the number of semi-speakers. Like many terms in linguistics (and elsewhere), the meanings and uses of speaker or semi-speaker can be subject to varied and sometimes conflicting interpretations, but they are ultimately necessary for a finer-grained approach to understanding a language's circumstances. For example, stating that a language has 100 native speakers is different from stating that a language has 100 semi-speakers. Note that while the term native speakers itself has been theorised and subject to critique (Davies Reference Davies1991; Singh Reference Singh1998; Bonfiglio Reference Bonfiglio2010), we do not risk inflating numbers by counting native speakers.
3. It is also important to address how language distinctions are made, in view of the regional, social, ethnolinguistic, cultural, and political implications that Grenoble mentions.
The ELCat FAQ explains that where sources may differ with respect to whether entities are considered different dialects of a single language or closely related languages, we strive to record the range of opinions involved on the relevant language pages.
4. It is also important to clarify how multilingualism is dealt with in ELCat.
ELCat includes information on what other languages are also used (when that information is available). For example, Baba Malay in multilingual Singapore and Malaysia, is spoken together with English, Malay, Mandarin, and Hokkien, among other languages (Lee Reference Lee2014).
5. Grenoble comments that language attitudes are not represented in ELCat figures.
Language attitudes are important, yet when constructing the index, we had to choose factors for which sufficient information is typically available, which unfortunately is not usually the case for attitudes. Nevertheless, attitude information is intertwined with and indirectly reflected in domains of use, speaker number trends, and transmission. Where sources do provide information on language attitudes, it is also included on the language pages, as in the example of Kinnauri in India, towards which older and middle-aged speakers have a strong positive attitude (Negi Reference Negi2012).
6. Grenoble also writes that it is misleading to compare government census data to information from a linguist's first-hand field report, as is comparing data across different years, or by different governments.
ELCat is subject to the vitality information available in sources, and the quality of available information can differ significantly from source to source. For example, there is typically a large difference in reliability between government census data and a linguist's first-hand field reports. For some languages we know little—Sentinelese is such an example. However, LEI's certainty scale mitigates this, qualifying the reliability based on the sources available. For areas where information is lacking or outdated, interested parties should be encouraged to take steps towards obtaining new and more accurate information.
7. Grenoble focuses in her response on the notion that enumeration and commodification are problematic.
ELCat is aware of these issues, and acknowledges that numbers in social investigations often lump categories, leaving some variables unattended. Yet, this is not an LEI-specific issue, but a general comment on the social sciences. When it comes to actual descriptions, ELCat prefers to have quantitative reports supplemented with qualitative information—in fact, this is what ELCat tries to do in the comments listed with the languages. To date, ELCat has not received feedback that its mention of numbers has caused distress in any community. On the contrary, enumeration can be a useful tool. ELCat has received requests from some communities to list their language as being even more highly endangered, saying that they want to use the numbers to encourage their community members to do more to protect their language, and that they would like to utilize these figures to ask for more resources to help the community combat language loss and to develop effective revitalization programs. As an example, the Administration for Native Americans, which is an indigenous-oriented agency in the United States, requires Native American groups to produce their own numbers to qualify for funding. ELCat takes the stance that there is a time, place, and need for numbers, and, when numbers are to be used, there should be well thought-out mechanisms for understanding them.
