Dictionary Society of North America
ABSTRACT

The Upper Nicola dialect of Nsyilxcn (aka Okaangan Salish) currently has fewer than twelve first-language speakers (twi-Sharon Lindley, p.c.). This paper describes the Upper Nicola Nsyilxcn Talking Dictionary Project, a community-driven, collaborative project that utilizes heritage language recordings and their associated transcriptions with the goal of developing an online dictionary resource for use by the Upper Nicola Syilx community. The content of the dictionary stems from recently digitized recordings (originally made by Dr. Yvonne Hébert between 1978 and 1980) of elder twi-Joseph Albert Michel2 reciting Nsyilxcn words and sentences along with their English translations. The collection now consists of 6,831 individual sound files, which our team is currently transcribing with the aid of published language materials. A significant number of lexical items are either unique to this dialect, or otherwise previously undocumented. Our team currently consists of a linguist and two Syilx Indigenous language learners from the Upper Nicola who are enrolled in University of British Columbia, Okanagan's Bachelor of Nsyilxcn Language Fluency Program. This project is important since there has never before been any published dictionary of the Upper Nicola dialect. This project also builds capacity within the community for language maintenance and provides a model for future collaborations between Indigenous communities and academic institutions.

Keywords

Relational Lexicography, Revitalization Lexicography, community-driven research, indigenous languages of North America, Colville-Okanagan, Salish

[End Page 107]

DEDICATION TO JOSEPH ALBERT MICHEL (BY TWI-SHARON LINDLEY)

To UNCLE! For he gave us skill, stamina and the knowledge of our past.

To UNCLE! For he gave us the love for life and taught us to respect it.

To UNCLE! Because he shared our burden when it threatened to slow our pace and kept by our side when we traveled lightly.

To UNCLE! Because in his eyes we have seen ourselves.

To UNCLE! [End Page 108]

INTRODUCTION

This paper describes an in-progress lexicographic project, the Upper Nicola Nsyilxcn Talking Dictionary, a collaboration between Syilx Indigenous language learners in British Columbia, the Upper Nicola Band, a linguist, and the Bachelor of Nsyilxcn Language Fluency degree program at the University of British Columbia, Okanagan (UBC Okanagan). Some forty-five years since its foundations were laid, this project aims to construct an audio dictionary of nearly 7,000 words and sentences stemming from archival recordings of Nsyilxcn (aka Colville-Okanagan) spoken by fluent elder twi-Joseph Albert Michel in inline graphic (Quilchena, BC) and recorded by linguist Yvonne Hébert between 1978 and 1980 (Michel 1978–80). We envision the resulting online dictionary to be primarily a resource for Upper Nicola Syilx language learners, though we hope that Nsyilxcn learners of other dialects may also find it useful.

Speech technologies, including talking dictionaries, hold great potential in supporting communities of Indigenous language learners (Galla 2016; Cox, Boulianne, and Alam 2019). In this era of language reclamation and indigenization, Indigenous communities are now taking leadership roles in digital language resource creation, resulting in a necessary paradigm-shift in how Indigenous dictionary and language technology work proceeds (Bird 2020). Our project is an example of relational lexicography, which Schreyer and Turin (n.d.) (see also Trotter et al. 2021) define as "a shift towards dictionaries that are created by and with speakers and learners of under-resourced languages themselves, which recognize the relationships in community between speakers, between dialects, and also between community members and academics." Because our methodology approaches dictionary-making as a language-educational experience, our project also exemplifies revitalization lexicography, which "uses dictionary-making processes to reverse language shift" (Anderson 2020, 7). Moreover, since our project is community-led, and the ownership of the product rests with the community, it follows the tenets of community-driven research (see Anderson 2020, 20). In particular, language learners from the Upper Nicola who are involved in this project:

  • ■ are enrolled in a community-based, academic language revitalization program, [End Page 109]

  • ■ work directly with heritage sound recordings,

  • ■ build their implicit understanding of the phonology, morphology, syntax and lexicon of Nsyilxcn, and

  • ■ make design decisions to ensure that the structure of the dictionary will address the practical needs of the Upper Nicola language community.

At the time of writing, the project is still very much in progress, and so here we focus on the project background, the people involved, and the relational methodologies we are employing rather than the content and structure of the dictionary itself. In describing our process as one that strives to maintain relational accountability (Wilson 2008) to the Upper Nicola community, we aim to contribute to the timely conversation about how Indigenous language communities initiate, guide, and respond to lexicographic work, and how Indigenous relational lexicography informs wider lexicographic theory and practice. Our project's methodology may serve as a model for other academic-based Indigenous language programs supporting community-driven lexicography.

We begin with background on the project and the people involved, followed by a description of some of the relational methodologies we have used and are currently using in the construction of this resource. We finish with a discussion of possible future directions as well as questions and concerns relating to this project.

BACKGROUND

Language and dialect information

Nsyilxcn (aka Colville-Okanagan; iso 639-3) is a Southern Interior Salish language spoken in south-central British Columbia and north-central Washington State by fewer than one hundred elderly speakers (First Peoples' Cultural Council [FPCC] 2022). The Upper Nicola dialect is spoken in the Nicola region of British Columbia (marked by a star in Figure 1) in an area adjacent to Nłeʔkepmx (Thompson) and Secwepemc (Shuswap) territories. [End Page 110]

Figure 1. Nsyilxcn (Okanagan) Speaking Lands (map by William Hay in )
Click for larger view
View full resolution
Figure 1.

Nsyilxcn (Okanagan) Speaking Lands (map by William Hay in Doak 1983, 17)

Intensive language revitalization efforts are now underway on both sides of the border, raising high hopes for a new generation of fluent speakers. These efforts include:

  • ■ the development of curriculum resources, for example, by the Paul Creek Language Association

  • ■ the establishment of a fluency-building program within a post-secondary institution: the Bachelor of Nsyilxcn Language Fluency (BNLF) at the University of British Columbia, Okanagan (UBC Okanagan) (Armstrong 2011)

  • ■ the establishment of community-level fluency programs, like the Salish School of Spokane, and [End Page 111]

  • ■ the language documentation and education initiatives of the En'owkin Centre in Penticton, British Columbia.

Motivation for an Upper Nicola Dictionary

A comprehensive dictionary of Colville-Okanagan was published by linguist Anthony Mattina in 1987 (Mattina 1987). That dictionary uses an orthography based on the Americanist Phonetic Alphabet (APA), a system originally developed by Euro-American anthropologists in the nineteenth century. The APA was designed to deviate minimally from the character set found on a typical keyboard and is widely used for writing North American Indigenous languages, notably Salish languages. An APA-based orthography was approved by a consortium of Okanagan bands in 1982. The main Nsyilxcn–English section of the dictionary is organized primarily by Nsyilxcn root, with larger stems interpolated throughout and example sentences to help illustrate word use. It includes material from a range of dialects, with particular emphasis on the Southern dialect areas where Mattina primarily worked. The more extensive Nsyilxcn–English side may be more di®cult to use for learners without a linguistic background who want to speak the language, however the less extensive English–Nsyilxcn index does not include additional contextual information, requiring users to perform a "double look-up" (Turin and Ferreira 2022), meaning that once the Nsyilxcn word has been identified in the English–Nsyilxcn index, the user must look up the word again in the Nsyilxcn–English side to see related forms and examples of usage.

An expanded online version of this dictionary, published and maintained by Anthony Mattina (n.d.), has since alleviated these di®culties and has been welcomed by students, teachers, and linguists alike. The online version includes improved indexing by root, as well as numerous other indexations, including part-of-speech and a range of semantic categories. Students and instructors in the BNLF program make regular use of this online resource, often noting lexical or phonetic differences between forms as found in the dictionary and their own ways of speaking.

The Upper Nicola dialect is, however, underrepresented in this dictionary, since Mattina did not work extensively with speakers from this area. This is important to consider, since the Upper Nicola dialect of Nsyilxcn differs from other dialects lexically (e.g., Upper Nicola uses [End Page 112] inline graphic rather than inline graphic for 'foot'; reportative kwukw is found only in the Upper Nicola) and syntactically, as evidenced by some differences in permissible word orders (Lyon 2013). These distinctive traits might be attributable to the influence of neighboring Nłeʔkepmxcín (Thompson) and Secwepemctsín (Shuswap) languages. Dialect differences are often a point of pride in the Upper Nicola language community and receive frequent mention. Efforts are underway within Upper Nicola to adapt immersion materials developed for other Nsyilxcn dialects into the Upper Nicola dialect. Creating dialect-specific curriculum and a dialect-specific dictionary could help legitimize dialect differences (Anderson 2020, 12) and elevate the status of the language in the community. Though there have been prior attempts within the Upper Nicola community to create a dictionary resource, no drafts of this work are available, and it may be lost (twi-Laura Manuel p.c. ca. 2010). The Upper Nicola community is currently creating a learner's dictionary, though this may include only "basic" vocabulary required for beginner-level speakers. These factors all speak to the need for an accessible and comprehensive Upper Nicola Nsyilxcn dictionary which can serve as benchmark resource for advancing language learners in the community.

Joseph Albert Michel

The late 1970s saw the first intensive language work by outside linguists in the Upper Nicola community. Dr. Yvonne Hébert, now deceased, conducted linguistic work in the Upper Nicola area between 1978 and 1980, working with fluent elders Joseph Albert Michel and Sharon Lindley, who have already been mentioned, as well as twi-Joe Pete Saddleman, twi-Johnny Archachan, twi-Nellie Guiterrez, twi-Louisa Roper, and twi-Rosie Tom. Hébert archived many of her materials at the Royal Museum of British Columbia in Victoria, though she also held a research contract with the National Museum of Man in Ottawa. After explaining her dissertation research (Hébert 1982) to Sharon Lindley (1953–2022), Hébert was introduced to Lindley's "Uncle" Joseph Albert Michel (Figure 2), a fluent speaker of Nsyilxcn, on the understanding that Lindley would receive copies of all the materials. Hébert's recording sessions with Joseph Albert provide the seed for our current dictionary project (Michel 1978–80).

Joseph Albert Michel (1908–1984) lived his entire life in inline graphic (Quilchena, BC). One of twelve siblings, he was raised in the language [End Page 113] by his mother Josephine Michel, who never spoke English. As evidenced by the recordings, he was at least trilingual in Nsyilxcn, Nłeʔkepmxcín, and English and possibly also spoke Secwepemctsín and Chinook Wawa. He worked during the day to provide irrigation for ranches in the area. Well-versed in traditional medicines, foods, and practices, he was a quiet man who enjoyed playing solitaire and listening to hockey games on the radio. Though he never had children of his own, he assumed the role of father to Kathleen Michel and her siblings. His interest in language, and his implicit knowledge of Nsyilxcn language structure and its vast lexicon, is clearly evident in Hébert's recordings.

Figure 2. Qwəymálst, "Uncle" Joseph Albert Michel, November 1978 (by Lisa Steigler in )
Click for larger view
View full resolution
Figure 2.

Qwəymálst, "Uncle" Joseph Albert Michel, November 1978 (by Lisa Steigler in Hébert 1982, 12)

Beginnings of the current project

In 2010 Sharon Lindley, the niece of the late Joseph Albert Michel, and a longtime language champion, obtained copies of the recordings from the Royal Museum.3 Lindley entrusted these to John Lyon, then a Ph.D. student in Linguistics at [End Page 114] University of British Columbia, Vancouver. Lyon then digitized them using the Linguistic Department's reel-to-reel equipment. In the previous year, Lyon had begun working with elders in the Upper Nicola at the behest of Lindley, who had requested assistance in documenting the language. With the permission of the Upper Nicola Band (UNB), Lyon built working relationships with Upper Nicola fluent elders twi-Lottie Lindley and Sarah McLeod, leading to Lyon's dissertation (2013), among other work.

The quality of Hébert's digitized recordings leaves much to be desired. First, different reels seem to have been recorded at slightly different speeds, affecting the tenor and quality of Joseph Albert's voice. Second, it is likely that some amount of tape degradation occurred during the roughly thirty years that the recordings were stored. Nevertheless, all recordings are of transcribable quality, and the historical value and extensiveness of the collection immediately presented itself as the seed for a dictionary project. Because many of the words in this collection have fallen out of general use, and the collection documents one of an older generation of speakers who did not speak English until adulthood, the value inherent in the collection arguably outweighs pursuing possible alternative approaches to creating a sound dictionary which might nevertheless fix the issue with sound quality, namely re-recording the items in the collection with one of our current fluent elders. In the summer of 2010, Lyon set to work dividing up the digitized thirty-minute reel sides into sound files, each consisting of an individual Nsyilxcn word or sentence as pronounced by Michel.

The Sound Files

The resulting collection contains 6,831 sound files of individual word and sentence pronunciations with translations from Michel.4 The sound files are on average roughly five seconds each for individual words and their translations, while sentence recordings may be as long as fifteen seconds. For each file, Michel typically pronounces the Nsyilxcn twice, followed by a spoken English translation. The translations are sometimes literal and sometimes paraphrastic.

Sound files were saved in WAV format and named according to Michel's English translation. The resulting list of file names shows that there are duplicate pronunciations of a significant number of the items, though we currently estimate perhaps 5,000 (73% of the total) are [End Page 115] unique entries. All duplicates are saved as separate files with an additional number in the file name. This allows the team (which currently consists of the authors) to later select the best quality pronunciations as entry defaults for the sound dictionary. Additionally, across duplicates of a given word, Michel sometimes provides different translations or adds additional information, all of which should be preserved. Files corresponding to English phrases or sentences, rather than words, are named after the most "salient" word, followed by the entire phrase or sentence enclosed in parentheses. For example, ʔupənkstspíntk 'ten years' is listed alphabetically under years (since spintk 'years' is the head of the compound) and given the file name "years_(ten_years)." This somewhat rough method allowed Lyon to categorize the contents of the recordings alphabetically. This organization of the files facilitates our transcription work, makes the contents searchable, and will eventually help with cross-referencing multiple English words to a particular sound file.

Hébert transcribed many of her recordings in a set of notebooks. A cursory check of these notebooks and Hébert's dissertation show that the entirety of Joseph Albert's recorded materials were not deposited at the Royal Museum and hence are not currently in our possession. We are endeavoring to locate additional recordings at the National Museum of Man in Ottawa, with whom Hébert held a research contract, or in the care of Hébert's relatives, in keeping with our goal that the dictionary include as many sound files, and be as comprehensive a reflection of the vocabulary of an Upper Nicola fluent elder, as possible.

Although the transcription work could in principle have begun in 2010, it was another nine years before that work began. As Kathleen Michel says, "the pieces first needed to be in place in the community."

RELATIONAL METHODOLOGIES AND LANGUAGE EDUCATION

As an example of relational lexicography, this project has rested on community support, expertise, and hard work since its inception. From the beginning, Sharon Lindley, who was concerned with the decline of language in her community, recognized the value of her Uncle's recordings. Lindley and Lyon together arrived at the idea of creating a "talking dictionary" resource for the Upper Nicola community. Lindley, knowing the potential language-learning opportunities this project might [End Page 116] bring to the community, always advocated strongly for a revitalization lexicographic approach: that speakers and learners in the community be directly involved as a language fluency-building opportunity. Below we discuss two such opportunities.

Transcription as a Group Classroom Exercise

Initial forays into dictionary transcription were done on the Upper Nicola reserve in collaboration with language students as part of the Nicola Valley Institute of Technology (NVIT) course INST 120 Nsyilxcn Transcription for Beginners, taught by Lyon in 2019. This course offers beginner-level training in Nsyilxcn phoneme recognition and production (oral and written), with particular focus on sounds not found in English (e.g., glottalized, uvular, and pharyngeal sounds), as well as practical training in using Mattina's (1987) print and online dictionaries. Seventeen students were enrolled, all from the Upper Nicola, ranging in age from sixteen to their late seventies. Several of the elder students were "silent speakers," defined as "those who have a good understanding of their language but do not speak it" (FPCC, n.d.), largely as a result of residential school-induced trauma (see Truth and Reconciliation Commission of Canada, 2015). Much of the in-class sound practice involved group repetition and performance as a way of protecting those who had traumatic associations with language and those who felt self-conscious about their pronunciation.

The task of transcribing Joseph Albert Michel's materials presented itself as a practical and grounded final-project opportunity for students to apply their developing literacy skills and their newly-refined abilities to use existing dictionaries in the creation of a new Upper Nicola dictionary. Sharon Lindley, one of the enrolled students, approved the use of her Uncle's materials in class. Students were enthusiastic and invested in the project, as our approach was relational, involving community members creating a dictionary from community materials to benefit that same community (Czaykowska-Higgins 2009).

Pairs of students signed up for an English letter corresponding to a folder of sound files alphabetized by English translation. Students then worked in pairs to transcribe sound files in class and at home, using Mattina's dictionaries to double-check the accuracy of their work. A student might first look up Michel's English translation in Mattina's English–Okanagan index. However, because the relevant entry in Mattina's [End Page 117] dictionary may be under an English synonym, or may not occur at all, students gained practice in identifying Nsyilxcn roots, prefixes, and suffixes. As a result, they were able to utilize Mattina's root and stem listings in the more extensive Okanagan–English section, thus increasing the possibility of a partial or full match. The importance of using Mattina's dictionaries as a point of reference in transcription cannot be overstated: because of the poor sound quality of many of the recordings, vowel quality is often obscured (e.g., [a] sounds similar to [ə]), glottal features on consonants are not always apparent (e.g., ejective [inline graphic] may sound more like plain [t]), and voiceless and sibilant sounds (e.g., [s, x, ł]) are often entirely obscured by background noise. Such sounds when obscured are nevertheless etymologically or morphologically reconstructable and often seem to emerge from the recording on second listening, after being independently identified. Such reconstruction is critical for maintaining linguistic accuracy.

The students' work collectively resulted in transcriptions for 300 entries of the talking dictionary. These were entered into a Google spreadsheet (Figure 3). Information by column includes ID number, the file name of the sound clip, and our working transcription, as well as columns for notes, corrections, etymological information, and semantic class.

Figure 3. Working dictionary spreadsheet
Click for larger view
View full resolution
Figure 3.

Working dictionary spreadsheet

Transcription with BNLF students

The Bachelor of Nsyilxcn Language Fluency, originally envisioned by Jeannette Armstrong in her 2011 position paper, is the first accredited post-secondary bachelor's degree in Canada delivered in an Indigenous language. It is a four-year program, with the first two years taught in the community and accredited through the Nicola Valley Institute of Technology and the second two years offered at the University of British Columbia, Okanagan. BNLF students come from different age groups, have a wide range of language backgrounds, and have varying levels of Nsyilxcn fluency. [End Page 118] During their fourth year, students are required to complete a language internship with a community organization or language research initiative, as well as completing a capstone project which consists of creating a "proof of concept" for a new language resource or a significant contribution to an existing resource. The Upper Nicola dictionary project provides an ideal internship and capstone project opportunity for BNLF students from the Upper Nicola. Overall, this project provides a unique example of how relational lexicography can function as community-driven revitalization lexicography within an academic setting.

Our dictionary project received funding in 2022 from UBC as part of a Community-University Engagement Support (CUES) grant, with additional funds contributed by the Delmas Foundation. In addition to receiving academic credit, BNLF students, particularly those from the Upper Nicola, can now get financial support for their work as undergraduate researchers. Beginning in the summer of 2021, Kathleen Michel, Joseph Albert Michel's niece and one of the inaugural BNLF cohort, began working on dictionary transcription, first as part of her internship and later in her capstone project, for which she refined a transcribed subset of the dictionary. Michel and Lyon met for two hours weekly for several months, listening to and transcribing sound files. In January 2023, Justine Manuel, now in her fourth year as a BNLF student, joined the project as a research assistant after having been granted access to the materials through an agreement with Oliver Lindley, Sharon Lindley's son and current owner of the materials. The team envisions that other Syilx students from the Upper Nicola will join the project in the near future.

At the time of writing, the team has transcribed roughly 3,500 entries, representing approximately half of the files. Michel and Manuel complete the initial transcriptions, then the entire team works together to refine them. Since the files are alphabetized by English headword, semantically related terms tend to occur in clustered sequences (see Figure 3), and the same root or stem is encountered multiple times as the team moves down the list, facilitating identification of the forms. Additional notes are added into the spreadsheet in the following circumstances: if the file name needs to be corrected, if morphological or etymological information is particularly relevant to a form, if the form does not occur in Mattina's dictionaries, or if the transcription is unclear. Concerning this last point, some sound files remain elusive [End Page 119] either because of poor sound quality or because the morphological structure of the word cannot be accurately ascertained. The team plans to conduct follow-up consultations with fluent elders to provide their expert opinions.

The relational methodologies we employ are important to consider. Manuel's and Michel's research assistantships have provided them with valuable language-learning experience and mentorship, and a strong voice in constructing a language tool that should soon be usable by the Upper Nicola community. The family connection to Joseph Albert Michel's materials is a particularly powerful motivating force for Kathleen Michel, who stated:

It was hard when I first started, I was overcome with memories, to hear Uncle's voice after probably thirty something years after he passed, and then I can just close my eyes and see him sitting at the kitchen table talking. It was tough, to start, being in a leadership position after Sister [Sharon Lindley] passed. But it is really healing, listening to him talk: The more I listen, the more I came to accept that this is where I was meant to be. Being in the language at the En'owkin Centre, and seeing the importance of it and how lucky we are to have these materials. It was a hard year, but if Sister and Uncle were here what would they say? "That's where you were meant to be, that's part of your family history now, and you have to work with it." So it's who I am. I want to see this to the end, to be there when it is finalized. Just to think that my Uncle started it, Sister carried it on, now I'm carrying it on. To see it to the end project, and to think how much it will help our people. To see how easy [language] is, if you can just listen to it. To see how we can connect through language with family and friends. Once we get back to the land, we can help our kids learn again.

(June 22, 2023 interview)

DEVELOPING AND MAINTAINING THE PROJECT

In this section, we discuss important considerations relating to growing and maintaining our community relations with respect to this project as well as developing and maintaining the dictionary structure and software. [End Page 120]

Community Consultation

Pending continued funding, the team plans on involving as many BNLF students from the Upper Nicola as possible in the transcription work. Our goal is to help develop the proficiencies needed for maintaining the dictionary within the community into the foreseeable future and to build capacity in the community for other related language work (Ogilvie 2011, Anderson 2020). As the transcription work nears its completion, the next phase of the work will begin. The future directions this project takes depend on the needs and interests of the Upper Nicola community, and any considerations relating to design, software, and final products cannot proceed without community input and support. As Anderson (2020, 102) has pointed out, we need to ask how we can involve community in the decision-making process.

Dictionary Structure and Software

The team envisions the eventual online dictionary as minimally consisting of an Nsyilxcn–English wordlist with sound files and a reverse English–Nsyilxcn index, but there is much potential beyond this. Depending on the results of the team's community consultations and who the eventual users of the dictionary will be, morphological parsing may also be included, along with a root- and stem-based listing, example sentences, indexing by semantic domain, and other features currently in Mattina's online dictionary. The team also anticipates creating a print dictionary, since print dictionaries still serve in many language communities as authoritative resources and as tangible reminders of the importance of language.

The choice of software will depend on the structural features of the dictionary, which in turn depend on the needs and desires of the community. Schreyer and Turin (n.d.) include as part of their online Knowledgebase a review of available digital lexicography tools. The team currently considers the following as holding good potential for this project: FirstVoices, FLEx / Language Forge, Living Dictionaries, Miromaa, Mother Tongues, TLex. In order to find the best match, the pros and cons of each software will need to be discussed with the Upper Nicola Language Department, and other communities of interest. Once our software has been identified, we can then export the information from our spreadsheet.

Community Access

With the exception of their use in the INST 120 course at UBC Okanagan, the Michel recordings have been largely kept within the Michel family. It is Oliver Lindley's intention, following [End Page 121] his mother's wish, to continue to limit access to the materials. The team hopes that eventually UNB members might be able to use them as a learning resource, while acknowledging Michel-family ownership of the materials.

Kathleen Michel prefers that her Uncle's materials constitute a stand-alone dictionary, separate from any other Upper Nicola language materials, though she is open to using a software platform which allows the UNB community to edit or update transcriptions and translations, such as FirstVoices, Living Dictionaries, Miromaa (see Schreyer and Turin n.d.). Justine Manuel (p.c.) sees language learners as the primary users of this dictionary, and while the interface should be as simple as possible, features such as indexing by morphological root would be of interest to the community.

Upper Nicola Band is open to hosting and administering our project, so we anticipate that the dictionary will be housed at an Upper Nicola server and website, an approach consistent with the principles of Indigenous data sovereignty (Schnarch 2004, Brinklow et al. 2019). However, Justine Manuel cautions against giving unmitigated access to all community members, since mis-transcriptions might occur. This concern is particularly pertinent given that there are two distinct, yet very similar, orthographies being used in Syilx communities. The o®cially approved APA-based orthography distinguishes plain resonant sounds [m, n, l, y, w] from a glottalized series [inline graphic, inline graphic, inline graphic, inline graphic, inline graphic]. A modified version of the original orthography developed by the Salish School of Spokane—and used in many learning programs—uses the glottalizes graphemes inline graphic, inline graphic, inline graphic, inline graphic, inline graphic to represent, ambiguously, either phonetic sequences of schwa plus plain resonant (i.e., [əm, ən, əl, əy, əw]) or the glottalized resonants themselves, which in effect conflates the two series of resonants. The potential for orthographic confusion in a dictionary with unmitigated access is a real concern, though some form of version control, or layered access, might offer a solution.

Other Challenges and Future Work

The place of our project with respect to other lexicographic work currently happening in the Upper Nicola remains unclear, and there are also questions relating to maintaining and growing our project. [End Page 122]

Regarding other lexicographic work, the Upper Nicola Band is currently developing a separate, learner's dictionary. It will be interesting to see whether this spurs additional lexicographic work with fluent elders in the community, and if, when these efforts come to fruition, Joseph Albert Michel's materials may potentially complement or be included within a larger project. As a dictionary representing the lexical and grammatical repertoire of a single fluent speaker, any data limitations arising from Michel's age, gender or, more generally, from language change or intra-dialectal variation, might be mitigated by contextualizing the Michel collection as a separate but important piece of a larger Upper Nicola dictionary which could include recordings from a demographically diverse set of speakers. Any steps taken in this direction would require permission from Oliver Lindley, Kathleen Michel, and their families.

An additional challenge relates to the fact that only two of the BNLF's four years are based at UBC Okanagan: While recently graduated students such as Kathleen Michel have access to continued funding through the En'owkin Centre and CUES funding, it may be a challenge to involve other future BNLF graduates from the Upper Nicola who are not employed with En'owkin or who no longer have access to research assistantships. This is of particular importance since more experienced students mentor less experienced students, in terms of both the language and our methodology. Having part of the team based at Upper Nicola (rather than at UBC Okanagan) may facilitate graduated students in continuing to work on the project and would serve the long-term goal of hosting and maintaining the dictionary at Upper Nicola.

The team is optimistic that in the near future community consultations will help answer the following questions: What e-dictionary software should be used? Where will the dictionary be hosted? What structure should be used for the dictionary, what kinds of information should be included, and how will learners use the dictionary? Who will have access to the materials, and what measures will ensure security of the information? Should a print dictionary be created as well as an e-dictionary? Should the Michel dictionary be designed as part of a larger dictionary of the Upper Nicola dialect? [End Page 123]

CONCLUSION

This community-driven project aims to provide a comprehensive resource to Nsyilxcn learners by converting a seed collection of heritage recordings by elder Joseph Albert Michel into an accessible online "talking" dictionary. This resource will enable learners to build their vocabularies and tailor their developing Okanagan grammars to the specific nuances of the Upper Nicola dialect. We envision the final online dictionary to be primarily used by the Upper Nicola community, to be flexibly designed, and minimally containing transcriptions, translations, and sound files. As a relational lexicographic project, community direction, support, and involvement is fundamental. BNLF student researchers on this project contribute to language documentation and resource development, while gaining transferrable skills that could be used in future language documentation and revitalization work. The project builds and strengthens working collaborations between Indigenous individuals and organizations (Upper Nicola Band) and academic institutions (NVIT and UBC Okanagan) and may serve as a procedural and pedagogical model that an academic-based Indigenous language program can use in support of community-driven lexicography.

John Lyon
University of British Columbia, Okanagan
john.lyon@ubc.ca
ḱᵂaḱᵂíslaɁqn Justine Manuel
University of British Columbia, Okanagan / Upper Nicola Band
justinemanuel@live.ca
xᵂəstalqs Kathleen Michel
En'owkin Centre / Upper Nicola Band
kmaymic@gmail.com
John Lyon

John Lyon is an assistant professor in the University of British Columbia, Okanagan's Department of Community, Culture, and Global Studies. He teaches in the Bachelor of Nsyilxcn Language Fluency Program, the first program of its kind in Canada. His linguistic research focuses on narrative documentation, syntax, and semantics of Interior Salish languages, most recently with St'át'imcets (Northern Interior Salish) and Nsyilxcn (Southern Interior Salish). He has completed dictionary projects for Coeur d'Alene Salish: Lawrence Nicodemus' Coeur d'Alene Dictionary in Root Format and Lawrence Nicodemus' Snchitsu'umshtsn File Card Collection, and he is currently working with the Upper Nicola Syilx community to produce a sound dictionary of heritage Nsyilxcn materials.

ḱᵂaḱᵂíslaɁqn Justine Manuel

Justine Manuel is an advocate for learning and reclaiming the Nsyilxcn language. She was raised in inline graphic, the Upper Nicola Band of the Okanagan Nation. Justine is currently enrolled in the Bachelor of Nsyilxcn Fluency Degree, graduating in June 2024 from University of British Columbia, Okanagan. She believes that it is important to carry the Nsyilxcn language forward to maintain connection to the land, the life force, and the ancestors. Justine is honored to take part in the documentation, analysis and preservation required to provide access to language in the Upper Nicola dialect through the Dictionary project.

xᵂəstalqs Kathleen Michel

Kathleen Michel is a member of the Upper Nicola Indian Band, and one of the inaugural graduates of the Bachelor of Nsyilxcn Language Fluency Program at University of British Columbia, Okanagan. She is working with John Lyon and other students from the Upper Nicola to transcribe heritage sound recordings of her Uncle Joe Albert Michel, made over forty years ago, and to process these into a format which will culminate in a sound dictionary resource for the community.

REFERENCES

Anderson, Patricia M. 2020. Revitalization Lexicography: The Making of the New Tunica Dictionary. Tucson: University of Arizona Press.
Armstrong, Jeannette. 2011. Discussion Paper: A First Nation Language Fluency Degree. Manuscript prepared for First Nation Education Steering Committee and Indigenous Adult and Higher Learning Association.
Bird, Steven. 2020. Decolonising speech and language technology. In Proceedings of the 28th International Conference on Computational Linguistics, 3504–19. Barcelona: International Committee on Computational Linguistics.
Brinklow, Nathan T., Patrick Littell, Delaney Lothian, Aidan Pine, and Heather Souter. 2019. Indigenous language technologies and language reclamation in Canada. In Collection of research papers of the 1st International Conference on Language Technologies for All, 402–406.
Cox, Christopher, Gilles Boulianne, and Jahangir Alam. 2019. Taking aim at the "transcription bottleneck": Integrating speech technology into language documentation and conservation. http://hdl.handle.net/10125/44841. Presented at the 6th International Conference on Language Documentation and Conservation (ICLDC), Honolulu.
Czaykowska-Higgins, Ewa. 2009. Research models, community engagement, and linguistic fieldwork: Reflections on working within Canadian Indigenous communities. Language Documentation and Conservation 3(1): 182–215.
Doak, Ivy G. 1983. The 1908 Okanagan Word Lists of James Teit. Missoula: University of Montana Occasional Papers in Linguistics, vol. 3.
First Peoples' Cultural Council (FPCC). 2022. The Report on the Status of B.C. First Nations Languages. Brentwood Bay, BC.
———. n.d. Reclaiming My Language. https://fpcc.ca/program/reclaiming-my-language/ (accessed August 1, 2023).
Galla, Candace K. 2016. Indigenous language revitalization, promotion, and education: Function of digital technology. Computer Assisted Language Learning 29(7): 1137–51.
Hébert, Yvonne. 1982. Transitivity in Nicola Lake Okanagan. Ph.D. diss., University of British Columbia, Vancouver.
Lyon, John. 2013. Predication and equation in Okanagan Salish. Ph.D. diss., University of British Columbia, Vancouver.
———. 2015. Three Okanagan stories about priests. In Proceedings of the 50th Annual International Conference on Salish and Neighbouring Languages, 285–362. Vancouver: UBC Working Papers in Linguistics.
Mattina, Anthony. 1987. Colville-Okanagan Dictionary. Missoula: University of Montana Occasional Papers in Linguistics, vol. 5.
———. n.d. (in progress). Colville-Okanagan. https://www.meltr.org/CvDict/lexicon/index.htm (accessed September 6, 2023).
Michel, Joseph A. 1978–80. Upper Nicola Okanagan field recordings, recorded by Yvonne Hébert. Royal BC Museum Archive.
Ogilvie, Sarah. 2011. Linguistics, lexicography, and the revitalization of endangered languages. International Journal of Lexicography 24(4): 389–404.
Schnarch, Brian. 2004. Ownership, control, access, and possession (OCAP) or self-determination applied to research. Journal of Aboriginal Health 1(1): 80–95.
Schreyer, Christine, and Mark Turin. n.d. Relational Lexicography Knowledgebase. https://knowledgebase.arts.ubc.ca/ (accessed August 1, 2023).
Trotter, Bailey, Mark Turin, Christine Schreyer, Meryl Bishop, and Victoria Sear. 2021. An open access toolkit for collaborative, community informed dictionaries. Presented at the International Conference on Language Documentation and Conservation, Honolulu, HI.
Truth and Reconciliation Commission of Canada (TRC). 2015. Honouring the Truth, Reconciling for the Future: Summary of the Final Report of the Truth and Reconciliation Commission of Canada. Library & Archives Canada.
Turin, Mark, and Natália Oliveira Ferreira. 2022. Indigenous lexicography: A review of recent dictionaries and works relating to lexicography. International Journal of Lexicography 35: 378–95.
Wilson, Shawn. 2008. Research Is Ceremony: Indigenous Research Methods. Winnipeg, MB: Fernwood Publishing.

Footnotes

1. This article appears in Indigenous Lexicography, a special issue of Dictionaries: The Journal of the Dictionary Society of North America 44(2) (2023), edited by Christine Schreyer and Mark Turin. It is open access under a Creative Commons CC-BYNC-ND license (https://creativecommons.org/).

In the print version, all illustrations are rendered in grayscale. Any color illustrations can be found in the open-access online version at Project Muse: http://muse.jhu.edu/resolve/213

2. twi- is a prefix attached by custom to the names of individuals who have passed on. For this paper, we use twi- on first mention of an individual's name, and thereafter it is implied.

3. The contents of the recordings were of two types: elder recordings of storytelling events made at Nicola Bay (see Lyon 2015) and lists of Nsyilxcn words and sentences read and translated by Joseph Albert Michel. The second type comprise the majority of the archived material.

4. A very small number of sound entries were pronounced by elder Nellie Gutierrez.

Share