Documentation of discourse and cultural activities to advance scientific knowledge of an endangered tonal language

记录话语和文化活动，以增进对濒危声调语言的科学认识

• 批准号: 1761421
• 负责人: Jonathan Amith
• 金额: $ 17.49万
• 依托单位: Gettysburg College
• 依托单位国家: 美国
• 项目类别: Continuing Grant
• 财政年份: 2018
• 资助国家: 美国
• 起止时间: 2018-08-01 至 2025-01-31
• 项目状态: 未结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=1761421&HistoricalAwards=false
• 关键词: Documentation; discourse; cultural; activities; advance

The world's estimated 7,000 languages are disappearing at an increasingly rapid rate. Parallel to language loss is the disappearance of cultural and economic activities. The present project, part of a long-term multidisciplinary effort, will emphasize video documentation to enhance a language documentation project that focuses on highly endangered cultural activities particularly prone to loss. This has the potential to increase scientific understanding of the spatial and motion components of grammar and to generate a valuable corpus of the distribution of complex tonal properties in narrative and other grammatical contexts. With up to 21 tonal contrasts on a single disyllabic sequence, Yoloxochitl Mixtec (YM), the language documented in this project, has one of the most extensive tonal inventories in any reported language. This inventory that carries great functional weight both lexically (distinguishing words) and morphologically (marking both derivational and inflectional processes), providing opportunities to deepen scientific advances both in terms of YM tone and cross-linguistic studies of complex tone systems. Broader impacts include the training of a US undergraduate in video documentation of endangered indigenous cultures. Additionally, annotated videos will be disseminated through Internet venues (e.g., Vimeo and YouTube). The entire subtitled corpus will serve as an educational resource for US high school and college students, especially those from diasporic Mexican Indigenous communities here in the US. In addition, video documentation will enhance intercultural understanding and has significant educational value for the community and the general public. The language and cultural and economic activities that will be documented are from Yoloxochitl, state of Guerrero, Mexico. The project continues the documentation of YM, a highly complex tonal language in which little comparative work has been done on this prosodic feature. The present project on YM will create the first extensive audio corpus of Mixtec lemmas (headwords). Video documentation will enable a focus on activities that are best documented and understood through audiovisual recordings. Visual imagery most effectively captures varied processes (e.g., preparation of plant material) and movements (e.g., the relationship between language and gesture; the documentation of manual skills) that would remain opaque in textual or simple audio descriptions. Moreover, the lexicon of a given language is connected to activities that must be conceptualized if the words themselves are to be understood. The expert team will record with two cameras (one stationary and one mobile) to allow the complete recording of each process while yielding footage for the creative editing of short vignettes on each activity. All footage as well as the final videos will be transcribed in YM, translated into English and Spanish, disseminated, and archived. It will serve as a primary source in an under-documented group of language for future comparative work on over three dozen recognized Mixtec languages while allowing acoustic analysis of the high number of tonal contrasts. Additionally, the focus on semantic domains involving ritual performance and material culture production will provide highly nuanced data of often archaic material in the linguistic and anthropological realm. The results of this project will be of value to linguists in a range of specialties (typologists, Oto-Mangueanists, historical linguists, tonologists, and semanticists) as well as to anthropologists studying Mesoamerican language and culture.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

世界上大约有7000种语言正在以越来越快的速度消失。与语言丧失并行的是文化和经济活动的消失。本项目是一项长期多学科努力的一部分，将强调录象记录，以加强语文记录项目，该项目的重点是特别容易丧失的高度濒危文化活动。这有可能增加对语法的空间和运动成分的科学理解，并产生一个有价值的语料库，用于描述叙事和其他语法语境中复杂音调属性的分布。该项目记录的Yoloxochitl Mixtec （YM）语言在所有已报道的语言中拥有最广泛的音调库，在一个双音节序列上有多达21个音调对比。这个清单在词汇（区分单词）和形态（标记衍生和屈折过程）方面都具有很大的功能权重，为YM音调和复杂音调系统的跨语言研究提供了深化科学进步的机会。更广泛的影响包括培训一名美国本科生拍摄濒危土著文化的视频。此外，注释视频将通过互联网场所（例如Vimeo和YouTube）传播。整个字幕语料库将作为美国高中生和大学生的教育资源，特别是那些来自美国散居的墨西哥原住民社区的学生。此外，视频记录将增进文化间的理解，对社区和公众具有重要的教育价值。将被记录的语言、文化和经济活动来自墨西哥Guerrero州的Yoloxochitl。该项目继续记录YM，这是一种高度复杂的调性语言，在这种韵律特征上几乎没有比较工作。目前在YM上的项目将创建第一个广泛的Mixtec引词（标题词）音频语料库。录像记录将使人们能够集中注意通过视听记录最好地记录和理解的活动。视觉图像最有效地捕捉各种过程（例如，植物材料的准备）和动作（例如，语言和手势之间的关系；手工技能的记录），这些在文本或简单的音频描述中仍然是不透明的。此外，一种给定语言的词汇与一些活动相关联，如果要理解单词本身，这些活动必须概念化。专家团队将使用两台相机（一台固定相机和一台移动相机）进行记录，以便完整记录每个过程，同时为每个活动的短视频的创意编辑提供镜头。所有的录像以及最终的视频都将以YM进行转录，翻译成英语和西班牙语，传播并存档。它将作为未充分记录的语言组的主要来源，用于未来对超过36种公认的Mixtec语言进行比较工作，同时允许对大量音调对比进行声学分析。此外，对涉及仪式表演和物质文化生产的语义域的关注将提供语言和人类学领域中通常是古老材料的高度细致的数据。该项目的研究结果将对不同专业的语言学家（类型学家、奥托-曼古语言学家、历史语言学家、语义学家和语义学家）以及研究中美洲语言和文化的人类学家都有价值。该奖项反映了美国国家科学基金会的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

项目成果

Leveraging End-to-End ASR for Endangered Language Documentation: An Empirical Study on Yolóxochitl Mixtec

• DOI: 10.18653/v1/2021.eacl-main.96
• 发表时间: 2021-01
• 作者: Jiatong Shi;Jiatong Shi. Jonathan D. Amith;Rey Castillo Garc'ia;Esteban Guadalupe Sierra;Kevin Duh

End-to-End Automatic Speech Recognition: Its Impact on the Workflow for Documenting Yoloxóchitl Mixtec

端到端自动语音识别：它对记录 Yoloxóchitl Mixtec 工作流程的影响

• 发表时间: 2021
• 期刊: First Workshop on NLP for Indigenous Languages of the Americas. 11 June 2021. https://www.aclweb.org/anthology/2021.americasnlp-1.8.pdf
• 作者: Amith, Jonathan D.;Shi, Jiatong;Castillo Garcia, Rey
• 通讯作者: Castillo Garcia, Rey