For easier reading, turn your device sideways.

The Library / Stories

Front cover of Spoken, Taught, Told and Remembered

Spoken, Taught, Told and Remembered

A Practical Guide for Tribal and Cultural Offices Building a Cultural Preservation Program

by Vikas Bakshi

Foreword

This is a practical guide for tribal cultural, language, historic preservation and related offices that want to build a preservation program their Nation owns and can keep running after the first grant ends. The figures in it come from public sources dated 2024 to 2026; please check each one against its source before quoting it in a grant application or briefing. Novels mentioned as examples are works of fiction published in an online library the author founded.

For the people in small offices who keep a Nation's memory.

Author's Note

I am a program manager by trade. Through most of my career, my job in any meeting has been to ask who owns a task, when it is due, what it will cost, and what happens when the person doing it goes on leave. I have run projects, cleaned up processes that had grown tangled over years, delivered systems that had to work on the first day, and connected tools that were never designed to talk to each other. None of that made me an expert on any Nation's culture. It did teach me how small offices succeed and how they fail.

I came to this work because I kept meeting people in tribal cultural, language and historic preservation offices who were doing three jobs at once. A language director told me she spent more hours on grant reports in one quarter than she spent with elders. A preservation officer I met at a conference kept the office's only copy of a decade of consultation records on a laptop that was older than his youngest child. One office I worked with had a cabinet of cassette tapes that nobody had played in twenty years because nobody was sure whose permission was needed to play them. These were not problems of commitment. Everyone I met cared deeply. They were problems of time, people and control, and those happen to be the problems a program manager is trained to work on.

I want to say plainly what I believe, because the rest of the book rests on it. In my own words: "A people's culture, meaning its language, stories, customs and ways of knowing, cannot be replaced, and once it fades, records alone will not bring it back. It has to be lived, and it has to be handed on." Recordings, files and archives matter, and much of this book is about them, but they are scaffolding. The building is people speaking, teaching, telling and remembering.

Storytelling has always been the way culture passes from one generation to the next, and today it has to compete for attention. It competes with screens, with long commutes, with jobs that take families far from home, and with the plain exhaustion of modern life. A Tribal office cannot solve that alone, especially with a small staff and limited resources. It needs systems that make the work lighter, partners who respect its authority, and funding that arrives in a form it can actually use.

I also believe, without reservation, that only a people owns its culture. Every Nation decides what is shared, with whom and when. A university does not decide that. A museum does not decide that. A vendor certainly does not decide that, and neither does a well-meaning volunteer or an outside author like me. The job of any tool, partner or helper is to make that ownership easier to exercise, not to take a piece of it in exchange for convenience. When I recommend a practice in this book, ask yourself whether it strengthens your Nation's control. If it does not, set it aside.

My measure of success for this work is simple to say and hard to reach: more of each Nation's culture spoken, taught, told and remembered by more of its own people, and understood and respected by more people beyond it, shared on the Nation's terms, for generations to come. The title of this book comes from that sentence. I return to it at the end, and I hope you hold me to it along the way.

A few words about what this book is not. Above all, I have not written a sales pitch. I founded a company that works in this field, and I know how easy it is for a guide like this to slide into advertising. I have tried hard to prevent that. Every tool recommendation in these pages is written as a set of criteria you can apply to any product or partner, with no brand names. If a recommendation cannot survive being stated without a brand, it does not belong here.

It is not a substitute for your Nation's own law, protocols and customs. Your council, your elders, your cultural committee and your own legal counsel have authority that no guidebook can override. Where I offer sample policy language, it is a starting point for your own review, nothing more. Where my suggestion conflicts with your Nation's way of doing things, your Nation's way wins.

And I am writing as an outsider to every tribe's office, not from inside one. I have worked alongside tribal offices, listened to many people, and read widely, but I am not a citizen of a Nation whose work I describe. I have kept every anecdote general so that no office, person or confidential matter can be identified. When I cite a named tribe, it is because that tribe's work has been described in public sources, and I cite those sources at the back of the book. No tribe mentioned here is presented as a client of any company, including mine.

Why write it at all, then? Because the gap I kept seeing was not a gap in passion or knowledge. Elders know what they know. Language teachers know how to teach. Archivists know how to care for a photograph. What small offices often lack is a simple, written way to organize the work so that it survives a staff change, a grant cycle, a flood or a funding cut. That is the part I can help with. Charters, decision charts, checklists, calendars, consent records and succession plans are not glamorous, and nobody goes into cultural work dreaming of a reporting calendar. But in my experience, the offices that keep their programs alive for twenty years are the ones that put a few dull, sturdy systems in place early and then let their people do the meaningful work on top of them.

I have tried to write for the person who has fifteen minutes between a council meeting and a home visit, who is the only staff member in the office, and who has been told to "start a preservation program" with a small grant and a deadline. If that is you, thank you for the work you do. I hope a few pages here save you a week somewhere, and that the week goes to an elder's kitchen table instead of a spreadsheet.

© 2026 CRUNOD LLC. All rights reserved.

Next chapter

How to Use This Book

This book is meant to be used, marked up and photocopied, not read once and shelved. Here is how I suggest approaching it.

Read Part One first. The Introduction and Chapters 1 through 3 lay the groundwork: why the timeline is short, how to scope a program you can actually run, and how to set consent and protocols before anything is recorded. Everything that follows assumes those foundations. After Part One, the chapters stand on their own, and you can take them in whatever order your office needs.

If you are a one- or two-person office, follow the small office path: Chapter 2 (a program charter), Chapter 3 (consent and protocols), Chapter 4 (recording elders now), Chapter 7 (governance and data sovereignty) and Chapter 12 (staffing and continuity). Those five chapters will let you begin recording safely, protect what you record, and make sure the work does not vanish if you leave.

If you are preparing to apply for funding, follow the grant-ready path: Chapter 2 (a charter gives a grant its scope), Chapter 10 (finding the money), Chapter 11 (managing a grant after the award) and Chapter 13 (measuring what matters). Funders want to see a clear plan, a credible capacity to manage money, and a way to show results. Those four chapters cover all three.

Every chapter ends with a short "Start Monday" paragraph that lists three things a one-person office can do in a single week. They are deliberately small. If you do nothing else with a chapter, do those three things.

Throughout the book you will find paragraphs that begin "As of 2026." Those hold time-sensitive details such as deadlines, dollar amounts and the status of particular programs. The main text describes programs by type so that it stays useful as specific programs change. Treat the "As of 2026" paragraphs as a snapshot, and check every one of them before you rely on it.

The templates and checklists are collected in Appendix A, and sample policy language is in Appendix B. Each chapter describes the tools it uses; the appendices give you the working versions. Please adapt them. Change the wording, add your own categories, remove what does not fit. Put any template that touches consent, ownership or access through your own council and legal review before you use it. Where a template uses an English term for a role or a category, consider whether your own language has a better word, and use it. A consent form written in a Nation's own terms carries an authority that a borrowed form does not.

A note on numbers. The figures in this book come from public sources dated 2024 through 2026, and I have used only figures that could be checked against a primary or authoritative source. Numbers change. Speaker counts are revised, appropriations move, programs open and close. Before you quote any figure from this book in a grant application, a council briefing or a press release, go to the source listed at the back of the book and confirm the current number. A funder who catches an out-of-date statistic will wonder what else is out of date.

A note on language. I use "Nation," "tribe" and "community" more or less interchangeably, and "office" to mean any tribal or cultural department doing this work, whether it is called a language program, a cultural resources department, a historic preservation office, a library, an archive, a museum or an education department. I use "citizens" for enrolled members, recognizing that some Nations prefer "members" or another term. I use "elders" in the broad sense of knowledge-holders, while knowing that in many communities the word carries a specific, earned meaning. Please read these words in whatever way fits your own Nation's usage.

Finally, the scope. This book is written for tribal offices in the United States, including Alaska Native villages. Native Hawaiian programs appear as examples because their language revival is the best-documented recovery story in the country, not because Native Hawaiian organizations face the same legal situation as federally recognized tribes. If you work in Canada or elsewhere, much of the practical advice will still apply, but the funding and legal details will not.

© 2026 CRUNOD LLC. All rights reserved.

Next chapter

Introduction: Time, People and Control

When I strip away the details, nearly every tribal cultural office I have encountered needs three things. It needs time, because the knowledge it protects is held by elders. It needs people, because its funding is small, short-term and competitive. And it needs control, because the recordings, files and records it creates can easily end up somewhere its own laws cannot reach. This book is organized around those three needs. Let me take them one at a time.

Time first. The people who hold a Nation's language, songs, place knowledge and family histories are, for the most part, old. Voice of America, reporting on Native languages in 2023, quoted Wilhelm Meya of The Language Conservancy saying that "the average speaker is about 75." (Meya also co-founded the Lakota Language Consortium, whose dispute with the Standing Rock Sioux Tribe I describe in Chapter 7.) Elders in Native communities also face shorter lives than other Americans. Final federal life tables put American Indian and Alaska Native life expectancy at 65.6 years in 2021, against 76.4 for the country as a whole. By 2024 it had recovered to 71.1, still 7.9 years below the national figure of 79.0. Put those two facts together, an average speaker near 75 and a life expectancy near 71, and you arrive at the central fact of this work: for many languages and many families, the window for recording and learning from first-language speakers is measured in years, not decades.

That is not a reason for despair. It is a reason for priorities. An office that knows its window is short can choose to spend its first dollars and hours on what cannot be recovered later. Paper records can usually wait a year. An elder's knowledge often cannot.

People next. The money available for this work is small, spread thin and delivered in pieces. The federal 10-Year National Plan on Native Language Revitalization, released in December 2024, counted about $41.5 million in federal Native language funding in fiscal year 2024. The same plan calls for roughly $1.5 billion a year. That means current funding is about 2.8 percent of what the federal government's own plan proposes. Most of what does exist is competitive, so offices spend scarce staff time writing applications that may not succeed, and when they win, they take on reporting requirements that cost roughly the same whether the award is large or small.

Historic preservation offices show the same pattern from a different angle. Tribal Historic Preservation Offices receive a formula grant from the National Park Service, and the number of eligible offices has grown faster than the money. The average award has drifted down to roughly $100,000 a year, and that money is expected to fund an entire compliance, survey and consultation program, often with one or two staff. I describe the details in Chapters 10 and 11. The point here is that most offices are not short of purpose or skill. They are short of hands and hours.

Then control. When an elder sits down to record a story, the result is a file. That file has to live somewhere, on a phone, a laptop, a hard drive, a university server or a cloud storage account. A tribe's legal authority is, for the most part, territorial: it applies on the Nation's lands and to its citizens. A file stored on a server in another state, owned by a company or a university, sits outside that authority unless a written agreement brings it back in. The same is true of a recording made by a visiting researcher, a set of photographs held by a museum, or a dataset gathered by an outside organization. The best-documented failure in this field in the United States is not a piece of software that stopped working. It is an ownership dispute in which elders' recordings, made on tribal land, ended up legally owned by someone else. Chapter 7 tells that story.

Control matters in a newer way too. Recordings of speech are now valuable raw material for automated systems that learn from recordings. A large collection of a language spoken aloud can be used to build products that the community never approved and may never benefit from. Tribal leaders and national organizations have asked for explicit tribal consent before any tribal data is used that way. The practical lesson for an office is that ownership, consent and limits on secondary use must be written down before anyone outside the Nation touches a recording, and that the cost of skipping that step is discovered years later, when it is hardest to fix.

There is one more fact that shapes everything else, and it concerns where the people are. According to the 2025 Harvard Project databook, which draws on the 2020 Census, about 18 percent of single-race American Indian and Alaska Native people in the Lower 48 lived on reservations. Put the other way, roughly four in five lived elsewhere: in cities, in towns near their home lands, or far away. Some of that dispersal was the direct result of federal policy, as Chapter 9 explains. Much of it reflects the ordinary movement of families for work, school and marriage.

This means a tribal office serves two very different audiences at once. The first is citizens away from home. They are usually well connected, with phones, broadband and social media, but they are often disengaged from the Nation's day-to-day life, not because they do not care but because the information does not reach them in time or in a form they can use. The second is elders at home. They are deeply engaged, they hold the knowledge, and they are often poorly connected. Broadband on tribal lands, especially rural tribal lands, still lags far behind the rest of the country, and many elders need someone to sit beside them and help with a device. An office that designs only for one of these audiences will miss most of its people. An office that designs for both has to think carefully about channels, formats and in-person help. I return to this in Chapter 9.

So what works? Looking across the programs that have lasted, from the Plateau Peoples' Web Portal in the Pacific Northwest to the Miami Tribe of Oklahoma's language and cultural center at Miami University, I see a five-part pattern. I will refer back to it throughout the book.

The first part is tribal ownership in writing. The Nation owns its materials, and a signed agreement says so. Partners hold materials as stewards, not owners.

The second part is tribe-set access rules. The Nation decides who may see, hear, copy or share each item, using categories that fit its own customs: public, citizens only, family only, elders only, restricted by season or by gender, and so on.

The third part is an institution that carries the technical maintenance. Running servers, applying security updates and migrating files to new formats are real and recurring jobs. Small offices rarely have the staff to do them. Programs that last tend to place that burden on a university, a regional consortium or another stable institution, under an agreement that keeps ownership and access decisions with the Nation.

The fourth part is regional peer training. Staff and community members learn from one another, across Nations, in workshops and networks. Training delivered once by an outside expert fades. Training that builds a circle of peers lasts.

The fifth part, and the most important, is an intensive human program at the center: master-apprentice pairs, language nests, immersion classrooms, regular elder visits. Technology works as the on-ramp. It brings people in, keeps them connected between sessions and holds the record. It does not, on its own, produce speakers or carry culture forward. People do that.

The rest of this book walks through how to build those five parts in a small office, one step at a time. Part One explains why now and where to start. Part Two covers the core work: recording elders, teaching language, digitizing collections, governing data, choosing partners and reaching citizens away from home. Part Three covers money and people: finding funding, managing grants and keeping staff and knowledge in place. Part Four covers proving results and staying alive after the first grant ends.

As of 2026: Several federal programs that matter to tribal offices were in flux. Digital equity grant funding had been cancelled in 2025 and was being reinstated in part after litigation; a large one-time grant program for Native entities' digital access had a deadline in November 2026; and the 10-Year National Plan, released in December 2024, had no public federal statement on its implementation as of October 2026. Chapter 10 gives the details. By the time you read this, all of these will have changed in some way, so check the current status before you plan around any of them.

Start Monday: 1. Write down, on one sheet of paper, the names of the five oldest knowledge-holders your office works with and the date of your last conversation with each. 2. Find out where every recording your office holds is physically stored, and who has the password to each location. 3. Ask your director or council liaison one question: "Do we have a written agreement with every outside partner that holds our materials?" Write down the answer, even if it is "I don't know."

© 2026 CRUNOD LLC. All rights reserved.

Next chapter

Part One · 1. The Closing Window

Every office I have worked with eventually asks the same question: how much time do we have? It is the right question, and it deserves an honest answer. The honest answer is that nobody knows precisely, because nobody keeps a complete count, but the evidence points clearly in one direction. For most Native languages in the United States, and for many bodies of family, place and ceremonial knowledge, the time left to learn directly from first-language speakers and lifelong knowledge-holders is measured in years.

I want to walk through the numbers carefully in this chapter, not to frighten anyone, but because an office that understands its own timeline can plan with clear eyes. I also want to show you how to read language statistics, because you will meet them in grant notices, news stories and funder reports, and they often disagree.

How many languages are there? The answer depends on who is counting and what they are counting. The federal 10-Year National Plan on Native Language Revitalization, released in December 2024, uses a figure of about 167 Native languages spoken in the United States today, citing a 2023 Bureau of Indian Affairs literature review. A 2016 joint federal analysis gave a range of 155 to about 175. First Nations Development Institute uses a figure of about 150 Native languages still alive in the United States. Voice of America reported in 2023 that about 100 remain in the United States, out of more than 500 spoken across the Americas in 1492.

Those numbers are not contradictions so much as different measurements. Some sources count only languages spoken within United States borders, while others count all of North America. Some treat closely related ways of speaking as separate languages, and others group them together as dialects of one language. Some count a language as "spoken" if a single elderly person can still converse in it, and others require a community of daily speakers. When you cite a number, name its source and, if you can, its scope. A grant reviewer will trust "about 167 languages, according to the 2024 federal plan" more than an unattributed round number.

The projections are sobering. The federal plan reports estimates that fewer than 20 Native languages will still be in use by 2050 without intensive intervention. Wilhelm Meya of The Language Conservancy has estimated that only 10 to 15 United States languages have viable prospects beyond 50 years. These are estimates, not certainties, and they depend heavily on what communities do in the next decade. They are also the reason this book exists.

The age profile is the most important number for planning. The federal plan cites a 2016 federal analysis of Linguistic Society of America data covering 194 North American languages. Of those, only 33 were spoken by adults and children. Another 34 were spoken by adults but by few children. Seventy-three were spoken almost entirely by adults over 50. Forty-nine were spoken by a few people, mostly over 70. Five may already have had no remaining speakers.

Read that list again from the point of view of an office. If your language is in the group spoken by adults and children, your challenge is to keep transmission going and strengthen it. If it is in the group spoken mostly by adults over 50, your challenge is to create new speakers among younger adults and children before the over-50 generation ages out, which gives you perhaps twenty or thirty years. If it is in the group spoken by a few people mostly over 70, your challenge is urgent: record everything you can, pair learners with every remaining speaker, and do it now. If your language has no remaining first-language speakers, your work is different but not hopeless, as I explain below.

Census data show the decline in homes. According to the Census Bureau, the number of people who spoke a Native North American language at home fell from 364,331 in the 2009 to 2013 survey period to 342,311 in the 2017 to 2021 period, a decline of about 6 percent in eight years. That count includes speakers of any proficiency, and it is heavily weighted toward a few large languages. Navajo alone accounted for 161,174 of those speakers, about 47 percent, down from 166,826 in the earlier period. The next largest, Cherokee, had 10,440 home speakers, followed by Zuni, Choctaw and Hopi, each with roughly 7,000 to 8,000. For most languages, the home-speaker count is far smaller, and the Census figure overstates their health.

Even large Nations count fluent first-language speakers in the low thousands or fewer. The Cherokee Nation is the largest tribe in the country by enrollment; it now counts 477,673 citizens. In January 2024 the Nation put its first-language fluent speakers at fewer than 2,000, most over the age of 70, and in November 2024 it honored about 1,500 speakers at its annual speakers' gathering. In 2016, the Chickasaw Nation reported fewer than 100 fluent speakers, all over 60, in a tribe of about 62,000, and its last monolingual speaker had died in December 2013. According to the then-acting commissioner of the federal Administration for Native Americans, as reported in 2021, the Kiowa had about 20 fluent speakers among roughly 12,000 members before the pandemic.

I share those examples because they are public, not because they are typical. Every Nation's situation is its own. Some small tribes have stronger home-language use than larger ones. Some languages with few first-language speakers have growing numbers of second-language learners who are now raising children in the language. A single statistic never tells the whole story.

There is no national speaker database. The federal plan itself names this as a barrier: no comprehensive database tracks speaker counts language by language across the country. That means your own office's count may be the best count that exists for your language. It also means that when you write a grant, you may need to explain how you arrived at your number. "Our language department maintains a list of first-language speakers, updated each spring through home visits and family reports" is a strong sentence. It shows a method, not just a figure.

The same gap exists for losses. No federal or peer-reviewed national count of first-language speakers lost each year has been published, and no national count of speakers lost to COVID-19 exists. I return to the pandemic in Chapter 4, because it changed the calculus for many offices.

How to read a language statistic. When you encounter a number about Native languages, I suggest asking five questions before you use it. First, who counted? A federal plan, a census, a nonprofit, a news reporter, a tribe's own department? Second, what did they count? Languages, speakers, home speakers, fluent speakers, first-language speakers, learners? These are very different things. Third, what is the scope? The United States, North America, one state, one Nation? Fourth, what year? A 2016 figure and a 2024 figure for the same language can differ for real reasons. Fifth, what threshold? Does "speaker" mean someone who grew up speaking the language, someone who can hold a conversation, or anyone who checked a box on a survey?

A good habit is to keep a small "language numbers" sheet in your files: one line per figure, with the number, the source, the date, the scope and the threshold. When you write a grant or a council briefing, pull from that sheet. When a number is updated, change it in one place. Appendix A includes a short guide you can copy.

The window, stated plainly. When I sit with a director and work through these figures for their own language, the conversation usually lands in one of three places.

For some offices, the window is a decade or two. The language is still spoken by middle-aged adults and some children. The task is to strengthen transmission, build teacher capacity and record the deepest knowledge-holders while they are able.

For many offices, the window is five to ten years. First-language speakers are mostly over 70. Every year without recording and pairing is a year of irreplaceable loss. The task is to prioritize elder recording above almost everything else, and to make sure every recording session also teaches a learner.

For some offices, the window has closed for first-language speakers, and the work now relies on recordings, documents and second-language learners. That is not the end. It is a different kind of beginning.

The hopeful turn. If you take only one fact from this chapter, take this one: languages can come back, and they have.

In 1983, a statewide survey in Hawaii found 37 Hawaiian-speaking children aged 18 and under. Hawaiian was, by any reasonable measure, close to the edge. Families, teachers and community leaders started language nests for very young children, then pushed for Hawaiian-medium schools so those children could continue their education in the language. By 2018, according to testimony before the Senate Committee on Indian Affairs, Hawaiian-medium education enrolled more than 3,000 students from preschool through grade 12. The demonstration school reported a 100 percent graduation rate since its first class in 1999, with more than 80 percent of graduates going straight to college, while 70 percent of its students were eligible for free or reduced-price lunch. Native Hawaiian organizations are not tribes, and their legal and funding situation differs, but the lesson about immersion holds everywhere: intensive, sustained use of a language by children and families produces speakers.

The Miami Tribe of Oklahoma offers another kind of hope. Its language was considered to have no speakers after the last one died more than 50 years ago. Working with Miami University, the Tribe has used archival documents and recordings to bring the language back into use among its citizens, and it is now being passed down again. A language without living first-language speakers is not necessarily lost if the records exist and the people are determined.

The Yuchi example, which I tell in Chapter 4, shows the value of even a short window used well. In the final months of the life of the last known first-language Yuchi speaker, learners visited her daily, and three prominent Yuchi speakers are credited to that time.

These stories share a pattern: the people who turned things around did not wait for perfect conditions. They started with what they had, they put learners next to speakers, and they kept going for years.

Taking stock: the "What We Hold Today" inventory. Before you can plan, you need to know what you have. The first tool in this book is a one-page inventory I call "What We Hold Today." It takes an afternoon to draft and a few weeks of conversations to complete. It has five sections.

Known speakers and knowledge-holders. List each person, with their approximate age, what they know (language fluency, songs, plant knowledge, place names, family histories, crafts, ceremonies to the extent appropriate), their health and willingness as far as you know, and the date of the office's last contact. Treat this list as confidential. Store it securely and share it only with staff who need it.

Existing recordings. List every recording the office knows about, whether it is held by the office, a family, a university, a museum, a past researcher or a radio station. Note the format (cassette, reel-to-reel, video tape, digital file), the approximate date, who is speaking if known, and who holds it. You will be surprised how many entries begin "a box in the back room" or "someone's uncle has tapes."

Collections. List photographs, documents, maps, objects and records the office holds or knows about, by rough category and quantity. You do not need item-level detail yet. "About 400 loose photographs, mostly 1950s to 1970s, unlabeled, in two boxes" is enough for now.

Classes and programs. List every current language class, culture camp, mentorship pairing or teaching activity, with who runs it, how often it meets, how many people attend and how it is funded.

Gaps and risks. Note what worries you most. A fragile collection. An elder whose health is failing. A partner who holds materials without an agreement. A single staff member who knows where everything is.

When the inventory is done, read it with your director, a trusted elder and, if possible, a council member. It will almost always point to one obvious first priority. Chapter 2 shows how to turn that priority into a program you can finish.

Start Monday: 1. Draft the "known speakers and knowledge-holders" section of the What We Hold Today inventory from memory, then mark each name with the date of your last contact. 2. Start a one-page language numbers sheet with every figure your office currently uses, noting source, year, scope and threshold for each. 3. Ask two elders or long-serving staff members where they know of recordings held outside the office, and write down every lead.

© 2026 CRUNOD LLC. All rights reserved.

Next chapter

2. Start With What You Have: A Program Charter for a Small Office

This is the chapter closest to my own working life, so I will begin with a confession. Early in my career, I watched a well-funded project collapse because everyone in the room agreed on the goal and nobody agreed on what the goal meant. Everyone wanted to "modernize the system." Some of us meant replacing it. Some meant patching it. Some meant training people to use it better. Two years and a great deal of money later, the project had done a little of each and finished none of them.

Cultural preservation programs are vulnerable to the same failure. Almost everyone in a Nation agrees that "we should preserve our culture." That sentence is a value, not a program. A program has a scope, an owner, a timeline, a budget and a way of knowing whether it worked. The work of this chapter is to turn the value into a program small enough to finish and sturdy enough to grow.

Why systems decide whether money can be used. I want to make one argument before getting into the tools, because it changes how you think about administrative work. When program staff across the country describe their barriers, many of the items on their lists are not linguistic or cultural at all. The federal 10-Year Plan names the shortage of teachers who are both fluent and trained, and the lack of facilities, as the largest constraints, and it calls competitive grants fragmented and short-term. First Nations Development Institute grantees added understaffing and turnover, limited access to translators and teaching expertise, elder health, curriculum gaps, poor broadband, and weak planning and assessment. Look closely at those lists and you will see grant reporting, assessment, recording and archiving, materials management and remote delivery. Those are systems problems.

The Association of Tribal Archives, Libraries, and Museums found the same thing from a different direction. Its 2024 reports, drawing on more than 700 survey respondents, found that two-thirds of tribal archives named conservation funds and technology as top needs, that more than 40 percent of tribal libraries could not meet their communities' needs, and that 60 percent of tribal museums wanted more staff. ATALM's preliminary 2026 findings, which the association notes are not representative, describe staffing in the field as "fragile, often solo" and identify unstable funding as the root cause of staff instability and program cuts.

Here is why that matters. A small office that wins a grant but has no system for tracking time, documenting match, filing reports and organizing files will spend so much of the award managing the award that little reaches the elders. A small office with simple, sturdy systems can absorb a grant, do the work and report on it without drowning. Your systems, more than your passion, decide whether scarce money can be used at all. That is not a cynical point. It is a protective one. Good systems protect the time you need for the meaningful work.

The one-page charter. A program charter is a single page that answers seven questions. I have used versions of it on projects of every size, and it works as well for a one-person language office as for a large technology rollout. Here are the questions.

Purpose. Why does this program exist? Write one or two sentences in plain language. "To record the knowledge of our remaining first-language speakers and make it available to citizens under rules set by the Nation" is a purpose. "To preserve our culture" is not specific enough.

Scope. What is included, and, just as important, what is not? "Included: audio and video recording of first-language speakers, transcription by trained learners, secure storage. Not included in this phase: digitizing photograph collections, public online access, curriculum development." Writing down what is out of scope is the single most useful thing a charter does. It gives you permission to say "not yet" to good ideas that would pull you off course.

Owner and decision-makers. Who is accountable for the program, and who decides what? I cover this in more detail below.

Timeline. When does the first phase begin and end? What are the key milestones? Keep it to three or four dates.

Resources. What money, staff time, equipment and space are available? Be honest. "One half-time coordinator, a $15,000 equipment budget, the use of the community center on Tuesdays" is a real resource list.

Measures of success. How will you know the first phase worked? Choose two or three things you can count, such as "twelve elders recorded, each session reviewed and approved by the speaker or family, each with a learner present." Chapter 13 goes deeper.

Risks. What could go wrong, and what will you do if it does? "Elder health: prioritize the oldest and most frail first. Staff turnover: keep a desk manual from week one. Equipment loss: back up every session the same day."

The charter should be signed or approved by whoever has authority over the program: a director, a cultural committee, a council. Approval matters because it turns the charter from your personal plan into the Nation's plan. When a new council member or a new director asks why you are doing what you are doing, the charter answers.

Deciding who decides. In my experience, more programs stall over unclear decision-making than over lack of money. Who approves a recording for release? Who decides whether a song can be shared with a partner? Who signs a data agreement? Who speaks for a family when the elder who was recorded has passed away?

Program managers use a simple tool for this called a RACI chart. The letters stand for Responsible (the person who does the work), Accountable (the one person who owns the outcome and signs off), Consulted (people whose input is required before a decision) and Informed (people who need to know after a decision is made). You list the key decisions down the left side and the people or roles across the top, then mark each box with R, A, C or I.

For a small cultural program, the key decisions usually include approving the program charter; choosing which elders to approach first; approving consent forms and protocols; approving a recording for any use beyond the office; setting access levels for each item; signing agreements with outside partners; approving grant applications; and approving public communications.

The roles might include the program coordinator, the director, the cultural committee or elders' council, the tribal council, the families of speakers, and legal counsel. Every Nation's chart will look different, and it should. In some Nations, a cultural committee holds authority that the council delegates to it. In others, the council keeps every decision about cultural material. In some, families or clans hold authority over particular songs or stories that no office can override.

The rule I insist on is this: every decision should have exactly one A. If two people are both "accountable" for approving a recording, then in practice neither is, and the recording will sit in limbo. If the honest answer is that a decision belongs to a group, name the group and name the person who records the group's decision.

The 90-day start plan. Once you have a charter and a decision chart, write a plan for the first 90 days. Ninety days is long enough to accomplish something real and short enough to keep momentum. Break it into three blocks of about a month each.

Days 1 to 30: set up. Get the charter approved. Draft consent forms and protocols (Chapter 3) and start them through review. Assemble a basic recording kit (Chapter 4). Set up file storage with a backup (Chapter 6). Draft the elder priority list. Start your desk manual (Chapter 12), even if it is only a few pages.

Days 31 to 60: first sessions. Conduct your first three to five recording sessions with elders who are willing and well enough. Pair a learner with each session. Review each recording with the speaker or family. Refine your process based on what you learn. Most offices discover in this month that sessions take longer, and elders tire sooner, than they expected.

Days 61 to 90: steady rhythm and review. Settle into a regular schedule. Report back to the director, the committee and the elders on what has been done. Revise the charter if needed. Decide what the next 90 days will hold.

A 90-day plan should fit on one page. If it does not, you are planning too much.

One pilot, not five. The most common mistake I see in small offices is starting five projects at once. A new coordinator arrives full of energy and begins digitizing photographs, recording elders, planning a language class, building a citizen newsletter and applying for three grants, all in the first month. Six months later, every project is a quarter done and none can be shown to a funder or a council.

I strongly recommend choosing a single pilot and finishing it before you start the next. For most offices, the right pilot is elder recording, for the reasons in Chapter 1: it is the work that cannot wait. For some offices, a different pilot makes more sense, perhaps because the elders are already well recorded, or because a fragile collection is at immediate risk, or because the council has asked for a particular result.

To choose, I use a simple worksheet. List your candidate pilots. For each, score from 1 to 5 on four questions. Urgency: will waiting a year cause permanent loss? Feasibility: can you do it with the staff, money and equipment you have now? Visibility: will completing it show the community and the council that the program works? Foundation: will doing it build systems (consent forms, storage, file naming) that later projects will use? Add the scores. The highest score is usually your pilot. If two are tied, choose the one that builds more foundation.

In the novel *The Wax Keeps the White*, one young woman sorting through her late grandmother's belongings is given a stack of two hundred shipping tags, and she chooses to label just nine objects, carefully, with their stories. That is fiction, but it captures the instinct I am recommending. Nine things done well, with their stories attached, are worth more than two hundred things half-labeled. A finished small pilot teaches you more, and earns more trust, than an ambitious plan in pieces.

When the office is one person. Everything above assumes you have at least a director and a coordinator. Many offices do not. If you are the entire office, the charter is still worth writing, perhaps more so. It is the document that tells your council what you can and cannot do with the time you have. It protects you from being asked to do everything. And it is the first thing your successor will read if you leave.

For a solo office, I suggest a few adjustments. Keep the charter to half a page. Make the decision chart short: list only the decisions that require someone other than you, and name who that is. Plan in 30-day blocks rather than 90. And find one ally, an elder, a retired teacher, a council member or a colleague in a neighboring Nation's office, whom you can call when you need a second opinion. A solo office is not a weak office if it has one good ally and a clear page of purpose.

As of 2026: The Association of Tribal Archives, Libraries, and Museums published preliminary findings in January 2026 from a survey conducted between December 2025 and January 2026. A fuller report may be available by the time you read this. Check ATALM's publications for the latest figures on staffing and needs in tribal cultural institutions.

Start Monday: 1. Write a draft one-page charter answering the seven questions, even if some answers are "to be decided." 2. List the five most important decisions your program will need to make and write down, honestly, who currently makes each one. 3. Score your top three candidate pilots on urgency, feasibility, visibility and foundation, and bring the result to your director.

© 2026 CRUNOD LLC. All rights reserved.

Next chapter

3. Consent and Protocols Before the First Recording

I once asked a room of cultural staff a simple question: "If you recorded an elder telling a story today, who would have the right to decide, ten years from now, whether that story could be played at a public event?" The answers ranged from "the elder" to "the family" to "the council" to "whoever has the file." Then someone in the back said quietly, "That's why we haven't recorded anything." The room laughed, but it was the truest thing said all day.

Fear of getting consent wrong keeps many offices from recording at all. The answer to that fear is not to skip consent, and it is not to wait for a perfect policy. It is to settle a few clear rules before the first recording, write them down, and improve them as you learn. This chapter shows how.

Why a signature is not enough. In mainstream research and media, consent is usually individual. A person signs a release form, and the recording can be used as the form describes. In many Native communities, knowledge does not belong to individuals in that way. Some stories are held by a family as a whole. Some songs are held by a clan or a society. Knowledge of a place, a ceremony or a medicine may be held in trust by particular people, who have no authority to give it away. An elder may be willing to share something that their family or community would not want shared outside.

The *Protocols for Native American Archival Materials*, which I describe below, make this point directly. For your office, the practical lesson is simple: an individual speaker's consent is necessary, but it is not always sufficient. Your consent process needs room for family, clan or community authority where your Nation's customs call for it.

That does not mean every recording needs a council vote. It means your consent form and your review process should ask the right questions. Is this knowledge yours to share, or does it belong to others? Is there anyone else who should be asked? Are there people who should never hear this, or times when it should not be played? The elder usually knows the answers. The form simply makes sure someone asks.

Sensitive categories. Not all knowledge is equally sensitive, and treating everything as top secret is as unhelpful as treating everything as public. The Protocols list categories of material that communities often consider sensitive. They include songs; ceremonies; healing and medicinal knowledge; oral and family histories; sacred sites and places; and genealogical information.

Your Nation will have its own list, and it may be quite different. Some Nations consider all songs public and some place names sensitive. Some restrict knowledge by season, so that certain stories are told only in winter, for example. Some restrict knowledge by gender, so that certain teachings are for women or for men. Some restrict knowledge by age or by initiation, so that certain material is for elders only.

Before you record, sit down with elders and your cultural committee and draft a short list of restricted or sensitive categories for your Nation. For each category, note who may hear or see the material, under what conditions, and who decides exceptions. This list becomes your restricted-content flag list. When a recording touches one of these categories, it is flagged at once and handled under the stricter rules.

Review before release. The second rule I recommend is that nothing leaves the office until it has been reviewed. Every recording should go through at least one review step before it is used for anything beyond the office's own secure storage. The review asks: Is the recording accurate? Did the speaker say anything they would want removed? Does it touch a restricted category? Who should be allowed to access it?

Ideally the speaker reviews their own recording, either by listening to it with a staff member or by reading a transcript. Where that is not possible, a family member or a designated cultural reviewer can do it. The review should be recorded in a simple log: date, reviewer, decision, any edits requested, the access level assigned.

Review takes time, and offices are tempted to skip it. Please do not. A review step is the moment when an elder's trust in the program is either confirmed or lost. Elders talk to one another. An elder who hears that their words were played somewhere they did not expect will tell others, and your next ten recording sessions will be harder to arrange.

Withdrawal of consent. People change their minds. An elder may decide, after hearing a recording, that they said too much. A family may decide, after an elder passes, that certain recordings should be restricted during a period of mourning. Your consent process should say clearly that consent can be withdrawn or narrowed at any time, explain how to do it, and explain what the office will do in response: restrict access, remove from any shared copies, or, if the person wishes, erase. Be honest about limits. If a recording has already been played at a public event, it cannot be unheard. But it can be removed from storage and from any further use.

What a consent form should contain. A good consent form for an elder recording program is short, clear and written in plain language. If your Nation's language has terms for these concepts, consider using them alongside English. The elements I recommend are these.

The speaker: name, and a contact person if the speaker prefers to be reached through family. Family or community sign-off: space for a family member, clan representative or cultural committee member to sign where your protocol calls for it. Purpose: what the recording is for, in a sentence or two. Uses allowed: for example, internal archive, citizen-only access, language classes, family copies. Uses excluded: for example, public online posting, commercial use, use by outside researchers, and any use to train any computer system without written tribal consent. Review step: a promise that the speaker or family will review the recording before any use beyond storage. Access level: a box to be filled in after review. Withdrawal: how to withdraw or change consent, and whom to contact. Signatures and date.

Notice the exclusion about training computer systems. Recordings of speech are valuable raw material for automated systems that learn from recordings. A large collection of a language spoken aloud can be fed into such systems to build products no one in the community approved. Naming that use explicitly, and requiring written tribal consent for it, closes a door that a general release form leaves open. I discuss this further in Chapter 7.

Appendix B contains sample consent language. Please read the disclaimer there and put any form through your own council and legal review.

The Protocols and the CARE Principles in office practice. Two sets of principles are worth knowing, both because they are good guidance and because funders and partners increasingly expect you to know them.

The *Protocols for Native American Archival Materials* were drafted in April 2006 at Northern Arizona University's Cline Library by 19 participants, including representatives of 15 Native American, First Nation and Aboriginal communities. They are addressed mainly to archives, libraries and museums that hold Native materials. They tell those institutions to consult with tribal leadership and cultural offices, record agreements in formal written memoranda, and let communities identify sensitive material. Where materials were collected in violation of a community's rights, the Protocols say community rights take precedence over open access. The Society of American Archivists declined to endorse the Protocols in 2008 and again in 2012, then endorsed them on August 13, 2018. That history matters. It shows how long it took for the mainstream archival profession to accept that communities have authority over materials about them.

The Protocols also make a point about copyright that every office should understand. Western copyright law protects authors and publishers. It generally does not protect the community depicted in a photograph or recording, and it does not address oral tradition or the cumulative, shared nature of traditional knowledge. A researcher who recorded an elder in 1970 may hold copyright in that recording under United States law, even though the knowledge in it belongs to the elder's people. This is why written agreements matter so much: they can establish rights that copyright law does not.

The CARE Principles for Indigenous Data Governance were drafted at a workshop in Gaborone, Botswana, in November 2018 and published in the *Data Science Journal* in 2020. CARE stands for Collective Benefit, Authority to Control, Responsibility and Ethics. The principles were written as a complement to the widely used FAIR principles, which say data should be Findable, Accessible, Interoperable and Reusable. FAIR is about making data technically easy to use. CARE adds people and purpose: data about Indigenous peoples should benefit those peoples, they should have authority to control it, those who use it have responsibilities, and its use should be ethical by the community's standards.

How do these translate into office practice? In my experience, they boil down to four habits. Ask before you act: consult the right people before recording, sharing or partnering. Write it down: every agreement, decision and restriction in writing, filed where your successor can find it. Label it: every item carries a record of who may use it and how. Review regularly: revisit access decisions as circumstances change, such as after a death, a change in leadership, or a new request.

Labels that travel with the file. One practical tool for the "label it" habit is the system of Traditional Knowledge Labels and Biocultural Labels developed by Local Contexts, a nonprofit initiative. The Labels are applied by communities themselves, through the Local Contexts Hub, and they attach a community's own rules to a digital item. Examples include labels indicating attribution (who should be credited), non-commercial use, community-verified content, or seasonal restrictions. The Labels are designed to travel with a file's metadata, so that an institution or researcher who encounters the item later sees the community's terms. The research data organization DataCite supports the Labels in research metadata.

Labels do not enforce anything by themselves. A label is a statement, not a lock. But labels make a community's expectations visible, which is often enough to change how a careful institution behaves, and they create a record that can support enforcement under a written agreement.

The warning pattern: consent for one purpose, reuse for another. The best-known United States case of misused Native research consent did not involve language recordings, but its pattern applies directly to cultural archives. Beginning in 1990, more than 200 members of the Havasupai Tribe gave blood for a diabetes study that had been agreed in 1989. The samples were later used, without the Tribe's knowledge, for research on schizophrenia, inbreeding and population migration. Some of the published findings contradicted the Tribe's own account of its origins. The Tribe and individual members sued. The case was brought against the Arizona Board of Regents and researchers affiliated with Arizona State University; some later coverage misattributes it to the University of Arizona. In April 2010 the dispute was settled, with $700,000 paid to 41 tribal members and formal apologies from the Board of Regents and the university.

Other Nations drew lessons from the case. The Navajo Nation has had a moratorium on genetic research since 2002, which it began reconsidering in 2017.

I tell the Havasupai story briefly because it shows how consent fails. The original consent was real. People agreed, in good faith, to a specific purpose. The harm came later, when the material was reused for purposes nobody had agreed to, by people who assumed that having the samples meant having the right to use them. Replace "blood samples" with "recordings of elders" and "genetic research" with "commercial voice products" or "a documentary," and you have the risk every cultural office faces. The protection is the same: name the allowed uses, name the excluded uses, require new consent for any new use, and keep the record.

A protocol is a community decision, not a form. In the novel *The Keeper of the Water*, a valley community decides together what to keep about its river and how to share it on its own terms. The book is fiction, but its central idea is one I want to underline. A protocol is not a document a coordinator drafts alone and files away. It is a decision a community makes, through whatever process the community uses, and the document is only the record of that decision.

Practically, that means your protocol should be drafted with elders and your cultural committee, discussed in whatever forum your Nation uses for such matters, and approved by whoever holds authority. It may take months. That is acceptable. While it is in progress, you can still record, under an interim rule: everything recorded is held at the most restricted level until the protocol is approved and the recording has been reviewed. The interim rule lets you begin while the community decides.

The protocol policy outline in Appendix A gives a structure: purpose and authority; definitions; sensitive and restricted categories; consent requirements; review and release process; access levels; withdrawal; partner and researcher requests; storage and security; periodic review of the protocol itself.

The sensitivity review workflow. Here is the sequence I recommend for every new recording, in order. First, recording is made under a signed consent form and held at the most restricted level. Second, the coordinator logs the recording with basic metadata and flags any known sensitive content. Third, the speaker or family reviews the recording, with a staff member present if helpful, and requests any edits or restrictions. Fourth, if flagged, a designated cultural reviewer or committee reviews it against the restricted-content list. Fifth, an access level is assigned and logged, with the reviewer's name and the date. Sixth, only then is the recording made available at its assigned level. Seventh, at least once a year, and after any death or significant change, access levels are revisited.

Seven steps sounds like a lot. In practice, for a simple recording of an elder telling a public story, steps four and seven take a few minutes. The workflow only becomes heavy when the content is sensitive, which is exactly when you want it to be heavy.

Start Monday: 1. Draft a list of five to ten categories of knowledge that might be sensitive or restricted in your Nation, and schedule a conversation with an elder or cultural committee member to review it. 2. Adopt an interim rule in writing: every new recording is held at the most restricted level until reviewed. 3. Compare your current consent or release form, if you have one, against the elements in this chapter and note which are missing, especially the exclusion for training computer systems.

© 2026 CRUNOD LLC. All rights reserved.

Next chapter

Part Two · 4. Recording Elders Now

If you read only one chapter of this book, read this one. Everything else in these pages can be done next year if it must be. Recording the knowledge of your oldest speakers and knowledge-holders cannot. This chapter is about how to do it well, safely and with respect, starting with very little equipment.

Why this is urgent. The elder population that holds this knowledge is small, poorer than average and has shorter lives than other Americans. According to the federal Administration for Community Living, there were 301,418 single-race, non-Hispanic American Indian and Alaska Native people aged 65 and older in 2019, or 702,325 when people of more than one race are included. The 2019 poverty rate among these elders was 18.7 percent, more than double the 8.9 percent for all older Americans. About half lived in seven states: Oklahoma, Arizona, California, New Mexico, North Carolina, Texas and Washington.

The pandemic made all of this worse, and fast. Provisional federal estimates showed American Indian and Alaska Native life expectancy falling 6.6 years between 2019 and 2021. Final data put it at 65.6 years in 2021, against 76.4 for the United States as a whole. By 2024 it had recovered to 71.1, still 7.9 years below the national 79.0.

The pandemic struck fluent speakers directly, and the losses were recorded tribe by tribe in public reporting. In 2021 the Cherokee Nation reported losing more than 50 fluent speakers. Early in the vaccine rollout, the Nation set aside doses for its first-language speakers, vaccinating them alongside frontline health workers from mid-December 2020. According to the then-acting commissioner of the Administration for Native Americans, the Kiowa lost two of their five fluent mentor speakers. Minnesota Public Radio reported that the Mille Lacs Band of Ojibwe lost about half a dozen of its roughly 25 fluent speakers in 2020; fifteen years earlier, the Band had 145. The Shawnee Tribe's language program, working with about 100 speakers, lost two elders to COVID-19 and a third at the age of 99.

Steve Premo, a Mille Lacs artist, made the same point to Minnesota Public Radio: when elders die, the stories they carry go with them and cannot be recovered.

There is no national count of these losses. No federal or peer-reviewed study has tallied how many first-language speakers or knowledge-holders died during the pandemic, and no national rate of speaker loss has been published. The losses I just listed are only the ones that made the news. Your office's own records may be the only account of what your Nation lost.

The Yuchi lesson. Here is the story I tell most often when an office asks whether recording is worth the effort. The last known first-language speaker of Yuchi, an elder in Oklahoma, was hospitalized with COVID-19 in November 2020. She survived. From then until she died, aged 96, in August 2021, learners came to her every day, talking with her and recording her. According to public reporting by PBS NewsHour, three prominent Yuchi speakers are credited to that extra time.

Notice what happened there. The goal was not to fill a hard drive. The goal was speakers. The recordings mattered, and they will matter for generations, but the deeper result was that three people learned to speak, in part because they spent those months at an elder's side. First Nations Development Institute, reviewing its language grantees, reached the same conclusion independently: record elders and document their knowledge now.

That is the principle of this chapter. The purpose of recording elders is speakers and relationships, not only files. Every session should produce a recording and also deepen someone's learning.

Setting priorities: the elder priority list. You cannot record everyone at once, so you need a way to decide who comes first. I suggest a simple list, reviewed with a trusted elder or cultural committee member, that considers four factors for each person.

Age. Older speakers generally come first.

Health. Elders who are frail or facing serious illness come first, though always with sensitivity. Some will want to record urgently; others will not want to be reminded of their health. Let them and their families lead.

Unique knowledge. An elder who is the last person to know a particular dialect, song cycle, place-name set or craft comes before one whose knowledge is shared by others.

Willingness. An elder who is eager to record comes before one who is hesitant. Never pressure a reluctant elder. Sometimes hesitation passes when they see others recorded respectfully. Sometimes it does not, and that is their right.

Rank each person on these factors, talk it over, and agree on a first group of five to ten. Keep the list confidential. It contains health information and personal judgments that should not be circulated.

The home visit. Most elder recording happens in elders' homes, and a home visit carries obligations that a studio session does not. A few practices I have learned from staff who do this work well follow.

Call ahead, and call again the day before. Confirm time, place and who will come. Ask whether the elder would like a family member present.

Bring something. Customs vary, but many communities expect a visitor to bring food, tobacco, a small gift or an honorarium. Ask your cultural committee what is appropriate, and budget for it. If your grant allows honoraria, document them carefully (see Chapter 11).

Arrive on time, and do not rush. The first part of a visit is often conversation, coffee, news about family. That is not wasted time. It is how trust is built, and the recording that follows will be better for it.

Explain the consent form clearly and without hurry. Read it aloud if helpful. Answer every question. If the elder wants time to think, or to talk it over with family, give it.

Keep sessions short. Elder fatigue is real. Sixty to ninety minutes, including conversation, is plenty for many elders. Some will want to go longer, and that is fine, but watch for signs of tiredness and offer to stop. It is far better to come back next week than to wear an elder out.

Close well. Thank the elder. Tell them what will happen next: when they will hear the recording for review, when you will come back. Leave contact information. Follow up as promised.

The learner in the room. My strongest recommendation is this: never send a staff member alone to record an elder if you can send a learner too. Pair a language learner, a young person, a family member or an apprentice with every recording session.

The learner's job is to listen, to ask questions, to practice, and, over time, to become someone the elder teaches directly. The learner may operate the equipment so the staff member can focus on the conversation. After the session, the learner can help transcribe or summarize, which is itself a powerful way to learn.

The pairing does three things. It turns every recording session into a teaching session. It builds a relationship between generations that may continue outside the program. And it means that when the elder passes, someone who sat with them is still here. The Yuchi story is the extreme case. Every office can do a smaller version.

A simple pairing plan lists each elder on the priority list and the learner or learners who will accompany sessions with that elder. Consider family ties (some elders prefer to teach their own relatives; some prefer not to), dialect, gender where customs call for it, and the learner's commitment. Ask each learner to commit to a minimum number of sessions, so the elder is not meeting a new face every time.

The recording kit. You do not need a studio. You need a few reliable tools and a habit of using them the same way every time. Here is a phone-first kit that a small office can assemble for a modest sum.

A recent smartphone, used only for program recordings, with plenty of storage. A phone that also holds a staff member's personal photos and messages is a privacy and security risk.

An external microphone, optional but strongly recommended. A small clip-on microphone or a compact directional microphone that plugs into the phone will improve sound quality enormously, especially in homes with televisions, refrigerators and children in the next room.

A small tripod or phone stand, so the phone is steady if you record video.

A portable battery pack and charging cables.

Headphones, so the operator can check sound during the session.

Printed consent forms, session metadata sheets and pens.

A dedicated storage device for transferring files the same day.

If you have more money, a dedicated audio recorder offers better sound and fewer distractions than a phone. But do not wait for better equipment. A phone recording made this month is worth more than a studio recording that never happens.

Audio or video? Both have value. Video captures gesture, expression and, for crafts and demonstrations, technique. Audio is less intimidating for many elders, produces smaller files and is easier to review. Ask the elder. Many offices record audio by default and video when there is a reason to.

Before the first real session, practice. Record a staff member or a willing relative in a typical home setting. Listen back. Adjust microphone placement. Learn the phone's recording settings. Equipment trouble during a session with an elder is avoidable and embarrassing.

The session metadata sheet. A recording without information about it is far less useful than one with it. Fill in a simple metadata sheet for every session, either on paper or on a form. Include the date and place; the speaker or speakers; others present, including the learner; the language or languages spoken; topics covered, in a few words; any restricted or sensitive content, flagged; the consent form reference; equipment used; file names; the name of the person who completed the sheet. Fill it in on the same day, while memory is fresh. After a few weeks, you will not remember who was in the room.

File naming. Choose a file-naming convention before your first session and never deviate from it. A good convention sorts well, is readable by a person and contains no spaces or special characters. One pattern I like is year-month-day, followed by a short speaker code, followed by a session number, followed by a file type: for example, 2026-03-14_SPK017_S02_audio. The speaker code links to your confidential speaker list, so the file name itself does not reveal who is speaking. Rename every file from the phone's default to your convention on the day you transfer it.

Same-day backup. On the day of every session, transfer the files from the phone to at least two other places, following the backup plan in Chapter 6. Do not leave the only copy of an elder's recording on a phone overnight. Phones are lost, broken and stolen every day.

Recording and verifying side by side. In the novel *Child of Ndar*, an old griot in Senegal asks her family for a recorder for the first time, and her son insists that every name she speaks be checked. The book is fiction, set far from any tribal office, but I find it a helpful picture of the balance this chapter asks for. Record generously, because time is short. Verify carefully, because a recording that contains errors, or that someone later disputes, loses its authority. The review step from Chapter 3 is where verification happens: the speaker or family listens, corrects, adds context, and approves.

Some offices also keep a "questions to follow up" list for each elder: names that need spelling, places that need locating on a map, stories that were begun but not finished. Bring the list to the next session.

When an elder passes. Despite every effort, some elders will pass before you reach them, and some will pass partway through. When that happens, the program's first obligation is to the family. Follow your Nation's customs about mourning, about saying the names of the deceased, and about playing their recordings. Some communities restrict recordings of a person for a period after death. Your protocol should say what to do, and your staff should know it before it happens.

Your second obligation is to the record. Note the date in your files. Review the access level of every recording of that person. Ask the family, when the time is right, whether they would like copies.

Your third obligation is to your staff. Recording elders means forming relationships with people who are near the end of their lives. Staff and learners will grieve. Chapter 12 has more on caring for them.

As of 2026: The Administration for Community Living's profile of older American Indians and Alaska Natives that I cite in this chapter was published in 2021 using 2019 data. The national life tables I cite were published in August 2026 and cover 2024. Check for newer editions before quoting these figures.

Start Monday: 1. Draft your elder priority list with five to ten names, ranked on age, health, unique knowledge and willingness, and review it privately with one trusted elder. 2. Assemble or order a basic recording kit, and practice one test recording in a real home setting. 3. Identify one learner who could accompany your first recording session and ask them to commit to at least four sessions.

© 2026 CRUNOD LLC. All rights reserved.

Next chapter

5. Language Classes and an Intensive Core

Nearly every tribal office I know runs, or wants to run, a weekly language class. It is the most visible language program a Nation can offer, and it does real good. It also cannot, on its own, produce new fluent speakers fast enough to replace the ones being lost. I want to explain why, honestly and with numbers, and then show how weekly classes can be part of a design that does produce speakers.

The hours arithmetic. Federal planners, in the 10-Year National Plan, estimate that producing a proficient speaker takes somewhere between 2,000 and 3,000 contact hours. Tribal consultation participants cited in the same plan put conversational ability at about 2,880 hours.

Now do the arithmetic for a weekly class. A class that meets for 90 minutes once a week, for 40 weeks a year, gives a learner 60 hours a year. At that rate, reaching 2,000 hours takes more than 33 years. Reaching 2,880 hours takes 48 years. A learner who starts in their twenties would be in their sixties or seventies before reaching conversational ability, if they never missed a session. Most of the remaining first-language speakers they could learn from would be gone long before then.

That arithmetic is not an argument against weekly classes. It is an argument about what weekly classes are for. They are an on-ramp: they introduce the language, build interest, connect learners to each other and to elders, and identify people who are ready for something more. The engine that produces speakers is something more intensive.

The intensive core. Three formats have shown they can produce speakers in a timeframe that fits a closing window.

Master-apprentice programs pair a fluent speaker (the master) with a committed adult learner (the apprentice) for many hours a week, often in daily life rather than a classroom: cooking, driving, working, visiting. The pair commits to using only the language during their time together. Over one to three years, a dedicated apprentice can reach real conversational ability.

Language nests immerse very young children, usually from infancy to about age five, in the language all day, often in a daycare-like setting staffed by fluent speakers. Children who attend a nest come out speaking.

Immersion schools continue that education through elementary and sometimes secondary school, with the language as the medium of instruction for most or all subjects.

The evidence for these formats is strongest for immersion, as the Hawaiian story in Chapter 1 shows: from 37 Hawaiian-speaking children in 1983 to more than 3,000 students in Hawaiian-medium education by 2018. The evidence for master-apprentice programs comes mostly from program reports. The Cherokee Nation's Master Apprentice Program has graduated nearly 60 adults, and its stated target is that graduates reach at least the intermediate-mid level on the proficiency scale used by the American Council on the Teaching of Foreign Languages (ACTFL).

The infrastructure is thin. According to the federal plan, some experts estimate that there are fewer than 50 active language nests in the United States, and there are about 40 K-12 immersion schools. Intensive programs are expensive. A tribal consultation participant cited in the plan described a Lakota immersion school that costs about $2.1 million a year, paid from grants and donations. Few offices can launch a school. Most can launch a master-apprentice pair.

Start with one pair. If your office has at least one fluent speaker willing to commit significant time and at least one learner willing to do the same, you can start a master-apprentice program. One pair is a program. Two pairs are a cohort.

A pairing agreement makes the commitment clear on both sides. I suggest it cover the following. The people: the speaker and the apprentice, and a program contact. The commitment: hours per week, for how many months, with a target total. Many programs aim for 10 to 20 hours a week for at least a year. The language rule: during program time, only the language is used, with agreed exceptions for emergencies or for clarifying a difficult point. Activities: what the pair will do together. Daily life is best. Payment: whether and how the speaker and apprentice are compensated. Paying both is common and fair. A fluent speaker's time is skilled work, and an apprentice who is paid can afford to give up other work. Records: how hours and progress will be logged. Support: what the office provides, such as transportation, materials, a mentor for the apprentice, check-ins. Ending or pausing: what happens if either person needs to stop. Consent: whether sessions may be recorded, under the consent rules in Chapter 3.

Pairs should be matched with care. Personality matters as much as fluency. An apprentice and a speaker who do not get along will not last a year. Consider family ties, dialect, gender customs and practical matters such as distance and schedules. Check in after the first month, and be ready to rematch.

Remote sessions can help. The Cherokee master-apprentice program moved online during the pandemic. Public reporting described dropped rural connections and full online days that proved too demanding; daily sessions were cut from eight hours to about six. Remote sessions are workable, but they need reliable connections and shorter days than in-person work. Chapter 9 covers remote readiness.

Teachers who are both fluent and trained. The federal plan identifies the shortage of teachers who are both fluent and trained to teach as one of the largest constraints on language programs. Fluent speakers are not automatically skilled teachers, and trained teachers are often not fluent. Building programs requires both.

For a small office, two practical steps help. First, offer fluent speakers who are willing to teach some basic support in teaching methods: how to plan a lesson, how to correct gently, how to keep learners speaking rather than listening. Many speakers are natural teachers and need only encouragement. Second, invest in your most committed learners as future teachers. The apprentice who finishes a master-apprentice program is the obvious candidate to teach the next weekly class, and eventually to become a master to a new apprentice.

Hawaiian language leaders, testifying before the Senate in 2018, made a point that applies everywhere: teachers cannot be imported. Parents must learn the language and become the teachers. A language program that depends on a small number of elder speakers and a few hired teachers will always be fragile. A program that turns parents into speakers creates homes where children grow up in the language, which is where every language ultimately lives or dies.

Weekly classes as on-ramps. With the intensive core in place, or at least planned, weekly classes take their proper role. A good weekly class introduces the language to newcomers, including citizens away from home who may never have heard it spoken. It builds community among learners who can support each other. It connects learners to elders, through guest visits and shared events. It identifies committed learners who are ready for master-apprentice pairing or other intensive study. And it maintains learning for people who cannot commit to intensive programs but want to keep the language in their lives.

In the novel *Clap on the Names*, a teacher has run a Saturday language and history class in a garage since 1971, and two generations of children have learned there and practiced on the drive home. It is fiction, but it pictures what a weekly class does best. It is steady, it is local, it belongs to the community, and over decades it shapes how families think of themselves. It does not, on its own, produce fluent speakers. It prepares the ground where fluent speakers can grow.

Apps and online tools in their place. Language apps, online video lessons and recorded courses are popular, and some Nations have built impressive ones. The Chickasaw Nation, whose language program was founded in 2007, launched an app in 2009 and runs an online television network, but pairs both with a master-apprentice program of several hours a day, five days a week. Its director, interviewed in 2016, put the balance well: "An app won't make you a proficient speaker, but it can help you learn the language."

That matches the evidence. No tribal language app has published download, retention or proficiency data, so nobody can say how many speakers any app has produced. Apps are useful as on-ramps, reinforcement and reference. They let a learner practice vocabulary on a bus or hear an elder's pronunciation late at night. Before building or buying one, be clear about its role and about who will own the recordings in it (see Chapters 7 and 8).

Records from day one. A language program that does not keep records cannot show what it has achieved, and so will struggle to keep its funding. Most programs count enrollment, which is easy. Far fewer count progress, which is what funders and councils increasingly want to know. Chapter 13 covers measurement in depth. Here I want to name the three records every language program should keep from its first day.

A class roster and attendance log: who enrolled, who attended each session, for how long. For master-apprentice pairs, log hours per week.

A learner progress record: a simple, regular assessment of each learner's ability. It need not be a formal test. A short conversation with a fluent speaker every few months, rated on a simple scale, is enough to show change over time. If your program has adopted a proficiency target, such as ACTFL levels, use that scale.

A session record: what was taught or practiced, by whom, with what materials. This becomes your curriculum over time, and your successor's guide.

Keep these records in one place, using the same format every time. A spreadsheet is fine. The point is consistency, so that a year from now you can say "our apprentices logged 2,400 hours between them and three moved up a proficiency level" instead of "we think it went well."

The hours-to-fluency worksheet. To help councils and funders understand the arithmetic, I use a one-page worksheet. List each program format you offer or plan to offer. For each, note hours per week, weeks per year, and hours per year per learner. Then divide 2,000 (or your preferred target) by hours per year to show years to proficiency. Put weekly class, weekend immersion camp, summer intensive and master-apprentice side by side. The visual comparison makes the case for intensive programs more clearly than any speech.

As of 2026: The federal 10-Year National Plan on Native Language Revitalization, which is the source of the hours and infrastructure figures in this chapter, was released in December 2024. As of October 2026, no public federal statement on its implementation had been found. The plan remains a useful reference for evidence regardless of its status.

Start Monday: 1. Fill in the hours-to-fluency worksheet for every language activity your office offers now. 2. Identify one fluent speaker and one committed learner who might form a first master-apprentice pair, and have an informal conversation with each. 3. Set up a class roster and attendance log for your current weekly class, if you do not have one, and record this week's session.

© 2026 CRUNOD LLC. All rights reserved.

Next chapter

6. Digitizing Collections and Records

Most tribal offices are sitting on more material than they realize. Photographs in shoeboxes. Cassette tapes in a filing cabinet. Video tapes from a culture camp in the 1990s. Council minutes going back decades. Program files from grants long closed. Letters, maps, newspaper clippings, handwritten word lists. Some of it is precious and fragile. Some of it is ordinary office paper. All of it is the Nation's record, and much of it is slowly deteriorating.

Digitizing that material can feel overwhelming. This chapter breaks it into steps a small office can take, starting with deciding what to digitize first.

What kinds of material. It helps to think in five broad groups, because each has different risks and needs.

Photographs and slides: prints, negatives, slides, albums. Paper photographs can last a long time in good conditions but fade, curl, stick together and suffer from moisture and handling. Color prints and slides from the 1960s through the 1990s are especially prone to fading.

Audio and video tapes: reel-to-reel tapes, cassettes, microcassettes, VHS and other video formats. These are the most urgent. Magnetic tape deteriorates chemically, and the machines needed to play older formats are becoming hard to find and harder to repair. A tape that plays today may not play in five years.

Documents: letters, manuscripts, word lists, notebooks, maps, newspapers. Paper is generally durable, but acidic paper yellows and crumbles, and anything stored in a damp place risks mold.

Council and program records: minutes, resolutions, correspondence, grant files, reports. These are part of the Nation's governmental record and often have legal significance. Many offices hold them only on paper or in scattered files on individual computers.

Born-digital material: files that were created digitally, such as digital photographs, recordings, emails and documents. These seem safe but are often the most at risk of all, because they live on a single laptop, a forgotten flash drive or an account whose password nobody remembers.

Setting priorities by risk, value and use. You will not digitize everything at once, so you need a way to decide what comes first. I use a simple priority matrix with three factors.

Risk: how likely is the item to be lost or damaged if it is not digitized soon? Magnetic tape, mold-affected paper, fading color prints and anything stored in a flood-prone place score high.

Value: how important is the item to the Nation? Recordings of first-language speakers, unique photographs of people and places, and founding documents score high. Duplicate copies of routine correspondence score low.

Use: how often will people want to see or hear it? Material that elders, teachers, families or researchers will ask for often scores high.

Score each group of material from 1 to 5 on each factor. Multiply or add the scores, as you prefer. The highest scores go first. In almost every office I have seen, audio and video tapes of speakers come out on top, because they score high on all three.

Condition triage. Before you digitize anything, look at its condition. A quick triage sheet helps. For each box or group, note: type of material; approximate quantity; storage location and conditions (temperature, humidity, light, any signs of water); visible damage (mold, brittleness, fading, sticking, broken cassette shells, tape shedding); and urgency (stable, monitor, act soon, act now). Material marked "act now" goes to the top of your list regardless of other scores.

Some damaged material needs a specialist. Moldy items should be isolated and handled with care, ideally by someone with conservation training, because mold can harm people as well as objects. Tapes that are sticky or shedding may need professional treatment before they can be played without destroying them. Do not try to play a damaged reel-to-reel tape on an old machine just to see what is on it. Find a professional digitization service, or a partner institution with the right equipment, and ask for advice first.

In the novel *The Line Around the Horse*, a teenager, after a 2008 flood, rinses photographs and freezes letters to save her grandmother's papers. The book is fiction, but its emergency steps reflect real conservation practice: wet paper and photographs can often be saved if they are handled quickly and correctly, and freezing wet documents buys time until they can be properly dried. It is worth knowing, before disaster strikes, which local institution has a freezer and a conservator you can call.

File formats for long-term keeping. When you digitize, you want files that will still open in twenty or fifty years. That means choosing widely used, open or well-documented formats, and keeping a high-quality master copy separate from any smaller copies made for everyday use. A simple guide follows, which you can adjust with advice from a partner archive.

For photographs and documents, keep master copies as uncompressed or losslessly compressed TIFF files at a resolution high enough to capture detail (many archives use 400 to 600 pixels per inch for photographs and 300 to 400 for documents). Make access copies as JPEG or PDF.

For audio, keep master copies as uncompressed WAV files (many archives use 24-bit, 96 kHz for analog transfers). Make access copies as MP3 or a similar compressed format.

For video, master formats are more complex and storage-hungry. If you lack specialist advice, a high-quality MP4 at the original resolution is a reasonable practical choice for a small office, with the original tapes kept in good storage if possible.

For text documents created digitally, PDF/A is a version of PDF designed for long-term keeping.

Never throw away the original physical item after digitizing unless your Nation has decided, through its own process, that it should be discarded. Digital files can be lost too.

Minimum metadata. A digital file without information about what it shows is almost useless. Every item you digitize should carry at least a minimum set of information, recorded in a spreadsheet or catalog and, where possible, embedded in the file itself. I recommend these fields at minimum: a unique identifier, following your file-naming convention; a title or short description; date, even approximate ("about 1965"); people shown or speaking, if known and if it is appropriate to record; place; creator (photographer, recorder, author); source (who held the original); rights and ownership; access level, following your protocol from Chapter 3; any Traditional Knowledge or Biocultural Labels applied; and the name of the person who catalogued it and the date.

Unknown information is normal. Many old photographs have no names or dates. Leave the field blank or mark it "unknown" rather than guessing. Then bring unidentified photographs to elders' gatherings and ask. Identification sessions are among the most rewarding parts of this work, and they often produce stories that are worth recording in their own right.

Seeing the Nation's knowledge beside the institution's. The Plateau Peoples' Web Portal, developed by Washington State University with partner tribes of the Columbia Plateau, offers a powerful model for metadata. For each item, the portal shows the holding institution's catalog record alongside a "Tribal Catalog Record" and a "Tribal Knowledge" section. The tribe's record can correct errors in the institution's description, add names and context, and give the item meaning in the tribe's own terms. Neither side can alter the other's record. The result is an honest picture: here is what the museum wrote, and here is what the people say.

Even if your office never builds anything like the portal, you can adopt its principle. When you receive a copy of material held by an outside institution, keep the institution's description, and add your own description beside it.

Attribution of every voice. The Ojibwe People's Dictionary, at the University of Minnesota, shows the value of careful attribution. Its 17,000 entries are built on elders' recorded voices. Each audio clip carries the speaker's initials and a link to a short biography, and an advisory council of Ojibwe speakers approves the content. A learner using the dictionary does not hear an anonymous voice. They hear a particular elder, from a particular place, whose life they can read about. That attribution honors the speakers and helps learners understand dialect and context.

Your office can do the same in its own catalog: every recording linked to a speaker record, with whatever biographical information the speaker or family wishes to share.

Access rules built into the archive. Some archive platforms designed for Indigenous communities let a community tag items by family, gender and place affiliation, limit sacred material to elders, and assign different rights to different kinds of users: community members, elders, outside visitors. One open-source platform of this kind grew out of work with an Aboriginal community in Australia, and by 2019 was in use by more than 600 groups. You do not need any particular product to apply the idea. Your access tiers (Chapter 7) should be recorded for every item, and whatever system you use should enforce them.

Digital return. Many tribes have important recordings and photographs held by outside institutions: universities, museums, libraries and government archives. Some of these were made generations ago by researchers who never returned copies. "Digital return" means making high-quality digital copies of those materials and returning them to the community, often with the community gaining control over how they are described and accessed.

A well-known example involves historical recordings of the Passamaquoddy Tribe. In 2019, Washington State University reported that Library of Congress engineers had digitized historical Passamaquoddy recordings and given the Tribe curatorial control, so that the Tribe curates them under its own customs.

To begin finding your Nation's materials in other institutions, write an outside-holder inquiry letter. Appendix A has a template. It should introduce your office and its authority; ask whether the institution holds materials relating to your Nation, its people, language or lands; ask for a list, with any existing descriptions; ask about the institution's policies on access, copies and community control; refer to the *Protocols for Native American Archival Materials*; and propose a conversation. Send it to the archives, library and museum departments of universities, historical societies and state archives in your region, and to any institution known to have worked with your Nation in the past.

The 3-2-1 backup rule. Here is the single most important technical rule in this book: every important file should exist in three or more copies, spread across two or more kinds of storage, with one or more of those copies kept somewhere physically separate.

For a small office, that might look like this: one copy on the office computer or a network drive; one copy on an external hard drive, updated weekly and kept in a locked cabinet; and one copy in secure cloud storage, or on a second external drive kept at another tribal building. Any of these can fail. The rule ensures that no single fire, flood, theft, hardware failure or deleted account destroys everything.

Write your backup plan down: what is backed up, where, how often, by whom, and how you check that it worked. Then test it. Once a quarter, pick a few random files and restore them from each backup. A backup you have never tested is a hope, not a plan.

Storage you can afford. Storage costs have fallen sharply, but they are not zero, and they recur every year. Digital audio and video masters take up a great deal of space. Before you commit to a storage approach, estimate how much material you will have in five years and what it will cost to store at that size, including backups. Ask partners whether they can provide storage under an agreement that keeps ownership with the Nation (Chapter 8). And remember that cloud storage, however convenient, places files on servers outside tribal jurisdiction, which is why the written terms in Chapter 7 matter.

As of 2026: The figure of more than 600 groups using an Indigenous community archive platform comes from a 2019 university report and has likely changed. Check current figures before citing them.

Start Monday: 1. Walk through your office's storage areas and fill in a condition triage sheet for every box or shelf of material, marking anything that needs action now. 2. Identify every audio or video tape in your holdings and list them by format and approximate date. 3. Write down your current backup arrangements honestly, and identify one change that would bring you closer to the 3-2-1 rule.

© 2026 CRUNOD LLC. All rights reserved.

Next chapter

7. Governance and Data Sovereignty

This chapter is about the question underneath every other chapter: who controls what your office creates? It can sound abstract, legal or technical. It is none of those things at heart. It is the question of whether a grandmother's voice, recorded in her kitchen, will still belong to her people in fifty years.

What data sovereignty means in practice. The Native Nations Institute at the University of Arizona, which announced the United States Indigenous Data Sovereignty Network in 2016, defines Indigenous data sovereignty as "the right of a nation to govern the collection, ownership, and application of its own data." The network's stated objectives include data storage and security and the treatment of data as intellectual property.

"Data" here means far more than spreadsheets. It includes recordings of elders, photographs, transcripts, maps of sacred and cultural places, genealogies, language materials, enrollment records, health information, survey results, and the records your office creates every day. Any information about the Nation, its people, its lands and its knowledge is the Nation's data.

Governing that data means deciding who may collect it, who owns it once it exists, who may see it, how it may be used, where it is stored, and what happens to it when a partnership ends. A Nation exercises data sovereignty every time it makes and enforces one of those decisions.

National tribal organizations have backed this principle formally. The National Congress of American Indians adopted Resolution KAN-18-011 at its 2018 Midyear Session, supporting United States Indigenous data sovereignty and the inclusion of tribes in the development of any broad data principles. In 2024 NCAI adopted Resolution NC-24-008, "Supporting Tribal Digital Sovereignty as an Exercise of Self-Determination."

Why written agreements are the main protection. Here is the hard legal fact at the center of this chapter. A tribe's jurisdiction is, for the most part, territorial. Tribal law applies on tribal lands and, in various ways, to tribal citizens. A recording stored on a server in another state, owned by a company or a university, is generally outside that jurisdiction. So is a copy held by a researcher who has moved to another country. So is a dataset used to build a commercial product.

Data does not stay where it was made. It is copied, uploaded, synchronized, backed up and shared, often automatically and often across borders. Cloud storage and collections used to train automated systems are not bound by tribal territory.

That means the main enforceable protection a Nation has over its data, once that data leaves tribal lands or tribal servers, is a written agreement. A contract, a memorandum of understanding, a data-sharing agreement, a research permit with terms, a license: these are documents that courts outside tribal jurisdiction can enforce. An unwritten understanding, however friendly, protects almost nothing when a dispute arises, when people change jobs, or when an organization is sold.

I cannot say this strongly enough. Before any outside party, whether a university, nonprofit, museum, vendor, volunteer, consultant or researcher, touches a recording, a photograph, a transcript or a record, get the terms in writing. The checklist later in this chapter tells you what the terms should cover.

The newest risk: automated systems that learn from recordings. Recordings of speech have become valuable raw material. Companies and researchers building automated systems that learn from recordings need large amounts of spoken language to train those systems. A collection of elders speaking a Native language is exactly the kind of material such systems can use, and once used, the knowledge in it cannot be pulled back out.

Michael Running Wolf, an Indigenous researcher in this field, warned in 2023 that there is "a huge commercial incentive to collect our language data," that such collection can "avoid the jurisdiction of a tribe," and that companies have no legal obligation to compensate communities for their data. Tribal leaders, NCAI among them in 2025, have warned that language data may be gathered from individual speakers without any tribal oversight at all.

In comments on federal policy dated March 15, 2025, NCAI asked for explicit tribal consent before tribal data is used to train such systems. It also asked for a bar on using administratively collected data, such as grant applications, for that purpose without consent; for civil and criminal penalties for scraping tribal data; for secure repositories for sensitive cultural information; and for support for tribal data centers. It cited Resolution NC-24-008.

At least one Nation has acted directly. On August 21, 2025, the Cherokee Nation's Principal Chief signed the Nation's first policy governing these automated tools. Under it, any such use involving the Cherokee language requires approval and input from fluent speakers, and a new governance committee will approve, prohibit or set limits on the tools. In announcing the policy, the Chief said Cherokee speakers must always lead every decision about such uses of the language.

For your office, the practical step is clear. Every agreement that involves recordings, transcripts or language materials should state that no party may use them to train any computer system, or allow anyone else to do so, without the Nation's prior written consent. Put the same exclusion in your consent forms (Chapter 3) and your volunteer agreements (Chapter 8).

The documented failure. The clearest public example in the United States of what happens without written control is the dispute between the Standing Rock Sioux Tribe and the Lakota Language Consortium, reported in detail by ICT News.

The consortium was founded by two non-Native people. It became the sole owner of a body of Lakota education materials, many of them recorded on tribal land with Standing Rock elders. In 2020, according to the reporting, the grandson of one of those elders was told her recordings could not be shared without a grant, and he later received a copyright notice for posting Lakota lessons. In May 2022 the Tribe banished the consortium from the reservation and required fair pay for anyone supplying language services. In October 2024, the tribal council approved legal action to obtain the recordings. The consortium has since said that all data will be "permanently stored and owned by the Tribes."

I tell this story not to condemn any individual. The people involved may well have believed they were helping. The lesson is structural. Elders' voices were recorded on tribal land, for the benefit of the people, and the legal ownership ended up somewhere else. The lock-in was not technical. No software failed. The lock-in was legal: copyright and ownership sat with an outside organization, and the Nation had no written agreement giving it control.

The cost of that kind of failure is discovered late, when it is hardest to fix. It surfaces on the day a family comes looking for their grandmother's recordings and learns that the rights sit with an outside group.

What happens when the record that defines you is held by someone else. In the novel *Tabios, With the S*, a typist's error in a 1946 federal file leaves the family of a 104-year-old veteran spending nine years proving that he exists. The novel is fiction and its subject is very different, but it captures a feeling many Native families know well: the frustration of depending on records you did not create and cannot correct. Data sovereignty is, in part, the work of making sure your Nation holds and governs the records that define its people, rather than relying on others to get them right.

Access tiers. Not everything a Nation holds should be equally available. Access tiers let you set different rules for different material. A typical set of tiers might be the following.

Public: may be shared with anyone, including online, in publications and with partners, subject to attribution.

Citizens: available to enrolled citizens of the Nation (and perhaps descendants or spouses, as your Nation decides), but not to the general public.

Family: available only to the family of the speaker or subject, as identified by the family.

Elders or knowledge-holders: available only to designated elders or people with particular standing.

Restricted: available only with specific permission, for example because of seasonal, gender, ceremonial or mourning restrictions.

Your Nation may want more or different tiers. Some Nations distinguish between clans or societies. Some add a tier for teachers, so that language instructors can use material that is not otherwise public. An access tier matrix, described in Appendix A, lists the tiers down one side and the kinds of action (view, listen, download, copy, share, publish) across the top, with a mark in each box showing what is allowed.

Assign a tier to every item during the review step in Chapter 3, and record it in your catalog. Whatever system you use to store and share material must be able to enforce your tiers. If it cannot, it is the wrong system for anything above the public tier.

Provenance. Provenance means the history of an item: who made it, when, where, under what agreement, who has held it since, and what changes have been made. Good provenance records make it possible to prove ownership, enforce agreements and resolve disputes. For every item, record who created it and under what consent or agreement; who has held it, with dates; whether it has been copied, edited or transcribed, and by whom; and what access tier it carries and who assigned it. Embed what you can in the file itself and keep the rest in your catalog. Provenance is dull work until the day a dispute arises, when it becomes the most valuable record you have.

Governing through policy and resolution. Day-to-day decisions belong with your office and your cultural committee, but the foundation should be set at the governmental level. A data governance policy, adopted by your council or by whatever body your Nation's law gives authority, establishes the rules that every office and partner must follow.

The outline I suggest covers: purpose and legal authority; definitions; ownership (all data about the Nation, its people and its knowledge is the Nation's property); collection (who may collect data and under what approval); consent (requirements for individuals, families and the community); access tiers; storage and security; sharing with outside parties, including required agreement terms; prohibited uses, including use to train any computer system without written tribal consent; retention and deletion; roles and responsibilities; complaints and enforcement; and review of the policy itself.

Before you bring a data governance policy, or any related agreement, to council, check a few things. Has legal counsel reviewed it? Has the cultural committee or elders' council been consulted? Does it conflict with any existing tribal law or policy? Is the resolution language clear about what is being approved? Who will be responsible for carrying it out? What will it cost? When will it be reviewed? Appendix A includes a council resolution checklist.

The written-terms checklist. Here is the core tool of this chapter. Any agreement with an outside party that will touch the Nation's data should include, at a minimum, these six terms.

Tribal ownership: the Nation owns all materials and data, including copies, transcripts, translations and derived works. The partner holds them as a steward.

No secondary use without written tribal consent: the partner may use materials only for the purposes stated in the agreement. Any other use, explicitly including use to train any computer system, requires the Nation's prior written consent.

Tribe-set access tiers: the partner must apply and enforce the Nation's access tiers.

Provenance metadata: the partner must keep and preserve provenance information, and must not strip or alter it.

Full export in open formats: at any time, and at the end of the agreement, the partner must provide all materials and metadata to the Nation in open, standard formats, at no or reasonable cost.

Verified deletion on exit: when the agreement ends, the partner must destroy all copies, including backups, within a stated period, and must certify in writing that it has done so.

If a partner will not agree to these six terms, think very carefully before proceeding.

The annual data audit. Once a year, review your data holdings and agreements. Do you know where every copy of your important materials is? Is every outside holder covered by a current written agreement? Have access tiers been reviewed, especially after deaths or significant events? Has anyone left the office who had access to systems, and has their access been removed? Have you tested your backups and your ability to export? The audit takes a day or two. Report the results to your director and, in summary, to council.

As of 2026: The Cherokee Nation's policy on automated tools was signed in August 2025; its full text was not found online at the time of writing. NCAI's comments on federal policy were submitted in March 2025. Federal policy in this area was changing quickly in 2025 and 2026, and new laws, rules or tribal resolutions may now apply. Check NCAI's current resolutions and guidance.

Start Monday: 1. List every outside party (university, museum, nonprofit, vendor, researcher, volunteer) that holds any of your Nation's materials, and mark which have a written agreement. 2. Compare one existing agreement against the six-term written-terms checklist and note what is missing. 3. Draft a first version of your access tiers, with a one-sentence description of each, and share it with your cultural committee for discussion.

© 2026 CRUNOD LLC. All rights reserved.

Next chapter

8. Choosing Partners and Tools

No tribal office does this work alone. Universities offer archives, linguists and students. Museums hold collections and conservation skills. Nonprofits offer funding, training and networks. Volunteers offer time. Vendors offer tools. Each can be a real help. Each can also, if the terms are wrong, take a piece of the Nation's control in exchange.

This chapter is about how to work with partners and choose tools without losing control. It draws on the models that have lasted and the failures that have been documented, and it ends with a scorecard you can use to evaluate any tool or partner without a brand name in sight.

Models that work. Four public examples show what good partnership looks like.

The Plateau Peoples' Web Portal is run by Washington State University's Center for Digital Scholarship and Curation with eight partner tribes of the Columbia Plateau. Tribal representatives select and curate the materials. A Tribal Advisory Board vetted the portal's Statement of Commitment. And, crucially, each tribe signed its agreement through its tribal government, not through an individual staff member or a department. That last detail matters. An agreement signed by a government binds the partnership to the Nation's authority, and it survives changes in staff on both sides.

The Miami Tribe of Oklahoma and Miami University show ownership done right. The Tribe's language had no living speakers for more than 50 years, and it is now being passed down again through work that combines archival research, community programs and a long partnership with the university. The Tribe holds the copyright to the archive software, the Indigenous Languages Digital Archive. When other tribes add their materials to the archive, those materials remain the property of each contributing tribe, while the university's center stores and maintains them. The center's National Breath of Life Archival Institute trained 141 tribal representatives from 65 language communities between 2011 and 2017. In 2025 a $2.1 million grant from the Mellon Foundation supported work with 20 tribal communities. That is the five-part model from the Introduction in action: tribal ownership in writing, tribe-set rules, an institution carrying the technical maintenance, and peer training across Nations.

The hub-and-spoke support model comes from an open-source archive platform built for Indigenous communities. Its developers found, as they reported in 2017, that the usual model for open-source projects, in which users contribute code and fixes, failed for tribal users who lacked "the resources, infrastructure, and programming skills." Their answer was a hybrid: a university maintains the core platform, regional hubs provide training and support, and tribal "spokes" set the features they need. The authors described community sustainability and platform sustainability as "inseparable."

A small example makes the same point at the scale of one teacher. A Wampanoag language teacher built a secure online learning site in which she approves every user by hand, as public radio reported in 2021. That gated access answered community fears about past exploitation and made people willing to put language online at all. You do not need a large institution to exercise control. Sometimes control is one person deciding, carefully, who comes in.

Why sustainability fails. When tribal digital projects fail, people often blame the software. In my experience, and in the public record, software is rarely the real cause. Projects fail because nobody owns the ongoing work. The grant that built the archive ends, the person who knew how it worked leaves, the server needs an update nobody can perform, the annual fee goes unpaid, and the system quietly stops. Or projects fail because ownership was never settled, as in the Standing Rock case in Chapter 7, and the Nation finds it cannot get its own materials back.

No United States tribal app or archive has published a public account of shutting down when its grant ended. That does not mean none have. It almost certainly reflects survivorship bias: projects that fail rarely write reports about it. When you hear about successful projects, remember that you are hearing only from the survivors.

The lesson is that when you choose a tool or a partner, the most important questions are not about features. They are about ownership, support and the exit.

Working with universities and researchers. Universities can bring a great deal: archives, linguists, students, grant-writing experience, storage and technical staff. They also have their own interests: publications, degrees, research funding and institutional reputation. Those interests are legitimate, but they are not the Nation's interests, and when the two conflict, a written agreement decides which prevails.

A memorandum of understanding (MOU) with an academic partner should cover the following: the parties, signed through the tribal government; purpose and scope; ownership of all materials, recordings, transcripts and data, which belongs to the Nation; the six written terms from Chapter 7; publication (whether the researcher may publish, and whether the Nation reviews drafts before publication); credit (how the Nation and individual speakers are credited); students (rules for student researchers and student-created material); storage (where materials are kept and how they are secured); reporting (how the partner reports to the Nation); term, renewal and termination; and dispute resolution. Appendix A includes an MOU outline.

Ask the university to identify one person who is accountable for the partnership on its side, and to tell you when that person changes. Academic staff move often. An MOU that survives a professor's departure is worth far more than a friendship that does not.

Working with museums and archives. Museums and archives that hold Native materials increasingly recognize their obligations to communities, in part because of the *Protocols for Native American Archival Materials*. When you approach one, refer to the Protocols, ask for an inventory of material relating to your Nation, and propose a written agreement covering access, description (including the Nation's own description, as in the Plateau model), copies and digital return. Be patient. Institutions move slowly. Be persistent. Many have moved further than you might expect.

Working with volunteers. Volunteers, including citizens, students, retirees and allies, can do a great deal of the practical work: scanning photographs, transcribing recordings, organizing files, helping elders with devices. They also handle sensitive material, often without the training staff receive.

Every volunteer should sign a short agreement. It should cover the role and tasks; confidentiality; ownership (anything the volunteer creates in the role belongs to the Nation); restrictions (no copies, no posting, no personal use, and no use to train any computer system); training required before starting; supervision; and ending the role, including returning materials and removing access. Write a short role description for each kind of volunteer task, so volunteers know what is expected and what is not. Appendix A has examples.

Working with vendors. Vendors sell tools: archive systems, storage, recording equipment, language apps, websites, databases. Many are honest and helpful. All of them are businesses, and their terms of service are written to protect the business, not the Nation.

Before signing up for any tool that will hold the Nation's materials, read the terms of service and privacy policy, or have someone read them for you. Look for answers to these questions. Who owns what you upload? Does the vendor claim any license to use your content, and for what? Can the vendor use your content to train computer systems, or share it with others who will? Where is the data stored? What happens if the vendor is sold, goes out of business or changes its terms? Can you export everything, in standard formats, at any time? What happens to your data when you stop paying?

If the standard terms do not protect the Nation, ask for changes. Some vendors will agree. If a vendor will not, consider another vendor.

A convenient offer. In the novel *The Artisan's Journal*, a young woodworker is offered a quicker, cheaper shortcut for the repairs her family's shop has always done by hand, and must weigh that offer against the reason the shop exists. The novel is fiction, but the decision at its center is one every office faces. A free tool, a generous partner offering to "take care of everything," a vendor with a slick demonstration: each can be the right choice. Each can also quietly change what the program is, who it serves and who controls it. Before accepting any convenient offer, return to your charter from Chapter 2 and ask whether the offer serves the purpose written there.

The vendor-neutral scorecard. Here is the tool I use to evaluate any system, platform or service that will hold or handle a Nation's materials. Score each criterion from 0 (not met) to 3 (fully met), and write a sentence of evidence for each score.

Tribal ownership in writing: the contract or terms state that the Nation owns all content and data.

Access controls: the system can enforce your access tiers, including family, elder and restricted levels, and can limit who sees what.

Full export: you can export all content and metadata at any time, in open, standard formats, without unreasonable cost.

Consent records: the system can store and display consent information and access decisions with each item.

Phone and low-bandwidth use: the system works on phones and on slow connections, for elders at home and citizens away.

Support and training: the vendor or partner provides training and timely support, and you know whom to call.

Cost after the grant: you know what the system will cost each year after any grant ends, and the Nation can afford it.

Exit plan: you know exactly what happens to your data if you leave, and the vendor will certify deletion.

A tool that scores 0 on ownership, export or exit should be rejected for anything above the public tier, however well it scores on everything else. Those three criteria are the ones that protect the Nation when something goes wrong.

The exit and export test. Before you commit to any tool, and once a year after, run an exit and export test. Pretend you are leaving. Request a full export of a sample of your materials, including metadata, consent records and access settings. Check that the files open in standard software on a computer that has never used the system. Check that the metadata is complete and readable. Note how long it took and what it cost. If the test fails, you have learned something important while you still have time to act. Chapter 14 makes this an annual drill.

As of 2026: The Mellon Foundation grant mentioned in this chapter was announced in 2025. Partnership programs, institutes and grant opportunities change often; contact the organizations directly for current offerings.

Start Monday: 1. Score the main tool or system your office uses to store recordings or photographs against the vendor-neutral scorecard, and note any criterion that scores 0. 2. Find the terms of service for that tool and look up the answers to the six vendor questions in this chapter. 3. If you have volunteers, check whether each has signed an agreement; if not, draft a one-page agreement this week.

© 2026 CRUNOD LLC. All rights reserved.

Next chapter

9. Reaching Citizens Away From Home

Most tribal offices were designed, in their habits if not their mission, to serve people who live nearby. Newsletters are printed and mailed. Classes are held in the community center. Events are announced on a bulletin board and by word of mouth. Those channels still matter, especially for elders. But they miss most of the Nation's people.

This chapter is about serving two audiences at once: citizens away from home, who are usually well connected but often disengaged, and elders at home, who are deeply engaged but often poorly connected. The two groups need almost opposite things. An office that serves both well has to plan deliberately.

Where the people are. According to the Harvard Project on American Indian Economic Development's 2025 databook, which draws on the 2020 Census, 510,162 single-race American Indian and Alaska Native people in the Lower 48 lived on reservations, out of 2,786,431 in total. That is about 18 percent, down from about 21 percent in 2010. Even counting all of the Census Bureau's designated Indian areas, only about 27 percent lived inside them. The Urban Indian Health Institute, citing the Census without a year, puts about 71 percent in urban areas.

These figures count people who identify as American Indian or Alaska Native, not enrolled citizens of any particular Nation, so your own Nation's pattern may differ. Many Nations know from their enrollment records that a majority of their citizens live away from tribal lands. The Cherokee Nation, for example, reported in 2026 that about 332,000 of its more than 477,000 citizens live outside its reservation.

People away from home are also undercounted. The 2020 Census missed an estimated 0.9 percent of American Indian and Alaska Native people outside tribal lands, and urban areas with more Native residents had lower response rates, according to an analysis by the National Council of Urban Indian Health.

How the dispersal happened. Part of this dispersal was the result of federal policy. The federal Relocation Program, which ran from 1952 to 1973, encouraged Native people to leave reservations for cities, offering job training and help with moving. Estimates of the number of participants range from under 70,000 to over 150,000. A 2024 study in the *Journal of Economic History* estimated that the program left roughly 100,000 more Native Americans living in its target cities, such as Chicago, Denver, Los Angeles and San Francisco, than would otherwise have been there.

Many relocated families lost daily contact with their communities and with the cultural practice that daily contact sustains. Their children and grandchildren, who make up much of today's urban Native population, often grew up far from their Nation's language and ceremonies, through no choice of their own.

I mention this history briefly, and without any blame on today's families. Citizens away from home are not less committed to their Nations. Many are deeply committed and hungry to connect. The barrier is usually distance, time and information.

What citizens away from home say they want. The best direct evidence is small and dated, but its findings are specific and useful. In 2014, the Native Nations Institute surveyed 74 enrolled citizens aged 18 to 29, from at least 38 tribes, who lived away from their tribal lands. The findings, published in 2015, included these: 31 percent reported no contact of any kind with their tribe during the previous twelve months. Sixty-three percent of those who answered lived more than three hours from home. Only 2 of the 74 had received employment notices from their tribe. Respondents said printed newsletters often arrived too late to be useful. In open-ended answers, they asked for things like a regularly updated website, monthly e-newsletters, social media, and emailed reminders to vote and invitations to events.

That is a pilot survey of young adults more than a decade ago, so treat it as a signal rather than a measurement. But the requests are practical and inexpensive. They come down to timely information, in the channels people already use, from the Nation itself.

Why informal groups are not enough. Where tribes do not provide such channels, communities build their own. Social media groups, group text threads and email lists connecting citizens in a particular city are common, and many do real good. But they have limits an office should understand.

They are not controlled by the Nation. A private social media group belongs, in practice, to whoever administers it and to the company that runs the platform. Content posted there is subject to the platform's terms, which may allow the company to use it in ways the Nation would never approve. Photographs and recordings shared in such groups can be copied and spread without any review.

They leave people out. Citizens who do not use that platform, or who are not invited, never see what is posted. Elders, in particular, are often missing.

They are fragile. When the administrator loses interest or the platform changes, the group and its history can vanish.

The answer is not to discourage informal groups. It is to give citizens official, Nation-controlled channels that are timely and useful, so that the informal groups can point to them.

What tribe-run online classes showed. The pandemic forced many language programs online, and the results were revealing. The Shawnee Tribe's online language sessions drew 15 to 20 learners, compared with 10 to 12 in person, and included members living in California and Virginia, according to Cronkite News in 2021. Public radio in Colorado reported the same year that Wampanoag, Acoma and Blackfeet language teachers reached citizens across several states and Canada through virtual classes, and that distant learners described a stronger sense of identity.

For many citizens away from home, an online language class was the first regular contact they had ever had with their Nation's language. That is a powerful on-ramp, and it costs relatively little to sustain.

The connection gap at home. Elders at home face a very different situation. Broadband on tribal lands has improved, but it still lags behind the rest of the country, especially in rural areas where many elders live.

The Government Accountability Office reported in 2022 that 18 percent of people on tribal lands lacked broadband in 2020, compared with 4 percent elsewhere, and called that a conservative estimate. The Federal Communications Commission's 2024 report, using end-of-2022 data at a higher speed benchmark of 100 megabits per second download and 20 upload, found about 24 percent of people on tribal lands unserved, compared with about 7 percent nationally. Rural federal reservations had only 46.1 percent access, and rural Alaska Native villages 44.2 percent.

Newer data show progress. The FCC's 2026 report, using data from June 2025, found 74.6 percent of people on tribal lands covered by 100/20 wireline service, and 87.6 percent when fixed wireless service is included. On rural tribal lands, wireline coverage was only 57.1 percent.

Speed matters too. The network testing firm Ookla found that 36 percent of 110 tribal nations had median fixed download speeds below 100 megabits per second, as reported in 2025. And the share of households in Native areas with no internet at all fell from 26.7 percent in 2017 to 15.3 percent in 2022, according to the Federal Reserve Bank of Minneapolis, still well above the national rate of about 9 percent.

Demand is not the problem. When the federal Affordable Connectivity Program paid up to $75 a month toward internet service on tribal lands, enrollment there ran 3.5 to 5 percentage points above the national rate, reaching 329,459 households, before the program ended in June 2024.

What is missing. Two pieces of evidence that would help every office do not exist. No published United States survey of digital skills among tribal elders could be found, and no published data on device ownership on tribal lands. Your office's own experience is, again, the best evidence you have.

The pandemic record shows what goes wrong without attention to elders' connections. As noted in Chapter 5, Cherokee master-apprentice participants who studied remotely faced dropped rural connections, and daily sessions were cut from eight hours to about six because full online days were too demanding. Teachers elsewhere reported a steep learning curve for elder speakers, a lack of devices, and paused programs.

Designing for both audiences. The practical conclusion is that every office-to-citizen tool must work on phones and on slow connections, and that elders need in-person help to use it. Here is a two-audience channel plan I recommend.

For citizens away from home: a regularly updated Nation website or citizen portal, with current information on programs, events, classes and elections; a monthly email newsletter, sent on a fixed schedule; online language classes, recorded where consent allows so people in other time zones can catch up; official social media accounts for key departments, linking back to the website; and email or text reminders for elections, enrollment deadlines and major events.

For elders at home: printed newsletters and mailed notices, sent early enough to be useful; phone calls from staff or volunteers for important announcements; radio, where a local station reaches the community; in-person classes and gatherings; and regular in-person tech-help sessions.

For both: the same information, at the same time, in every channel. An announcement that reaches citizens in Phoenix by email on Monday and elders at home by mail three weeks later tells the elders they come second.

In the novel *Good Evening, Home*, an eighty-year-old host brings back his Haitian Creole radio hour for listeners scattered from Miami to Brooklyn, and his grandson builds a segment from decades of old tapes. It is fiction, and a different community, but the idea is worth holding: one trusted, regular channel, in the people's own voice, can hold together a community scattered across a country. For some Nations that channel is a radio program. For others it is a monthly newsletter or a weekly online class. The format matters less than the trust and the regularity.

Low-bandwidth design. When you build or choose an online tool, apply a short checklist. Does it work on an older phone? Does it work on a slow connection, and can a page load without large images or videos? Can audio be played without video? Can material be downloaded for offline listening, where the access tier allows? Is the text large and clear? Are buttons easy to tap? Does it avoid requiring frequent password resets or complicated sign-in steps? Has an elder actually tried it, with you watching quietly?

That last question is the most important. Watch an elder use a tool for ten minutes without helping. You will learn more than from any specification.

Elder tech-help sessions. Elders at home often need someone to sit beside them while they learn to use a device or a tool. Family members are the most natural helpers, but not every elder has family nearby. Senior centers, tribal libraries and office staff can fill the gap.

A simple session plan: a regular, advertised time and place, such as an hour every other week at the senior center; one helper for every two or three elders; a focus on one task per session (joining an online class, listening to recordings, reading the newsletter online, video-calling a grandchild); written step-by-step instructions in large print to take home; patience above all; and a phone number to call between sessions.

Contact and consent for at-large citizens. To reach citizens away from home, you need their contact information, and you need their permission to use it. Many offices find their enrollment records have outdated addresses and no email or phone numbers at all.

An at-large citizen contact form, offered online and on paper at events, can collect name; enrollment or citizen identification where appropriate; mailing address, email and phone; preferred channels; interests (language classes, cultural events, elections, jobs, youth programs); and consent to be contacted, with a clear option to unsubscribe. Store this information securely under your data governance policy. It is the Nation's data.

A content calendar. Finally, plan your communications on a simple calendar. List each channel and what goes out, when. Monthly newsletter on the first Tuesday. Class reminders the day before each class. Election reminders on a fixed schedule before each election. Seasonal features. A content calendar keeps communications regular even when staff are busy, and it makes it easy to hand the work to someone else if needed.

As of 2026: The Affordable Connectivity Program ended in June 2024. Federal broadband and digital access programs for tribes were changing in 2025 and 2026, as described in Chapter 10. The FCC's 2026 broadband report cited here was released in August 2026 using June 2025 data; check for newer reports.

Start Monday: 1. Find out what share of your Nation's enrolled citizens live away from tribal lands, using enrollment records if your office can access them. 2. Ask three elders and three citizens living away from home how they currently hear about the Nation's cultural programs, and write down their answers. 3. Pick one upcoming announcement and send it out in every channel on the same day.

© 2026 CRUNOD LLC. All rights reserved.

Next chapter

Part Three · 10. Finding the Money

There is no shortage of advice telling tribal offices to "diversify funding" or "pursue grants." There is a shortage of plain description of what the funding landscape actually looks like for a small cultural office. This chapter tries to supply it, and then offers tools for deciding which opportunities are worth your limited time.

I have written this chapter to age well. The main text describes kinds of funding. Specific programs, amounts and deadlines appear in "As of 2026" paragraphs, because they will change. Please check every one of them before you plan around it.

The shape of the landscape. Funding for tribal cultural and language work has a recognizable shape: a few large, one-time pots of money, and many small, competitive, short-term ones. Recurring, predictable funding is rare.

The scale of the gap is stark. The federal 10-Year National Plan on Native Language Revitalization counted about $41.5 million in federal Native language funding in fiscal year 2024, against the roughly $1.5 billion a year the plan calls for. That is about 2.8 percent. The plan proposes $16.7 billion over 2025 to 2035, including $6.86 billion in direct, non-competitive funding to all 574 federally recognized tribes and Native Hawaiian organizations and $2.88 billion to train 10,000 teachers, with a goal of increasing proficient speakers by 25 percent before 2036. The plan also offers a historical comparison: from 2006 to 2024, federal spending on language revitalization was less than one cent for every dollar of the at least $23.3 billion, adjusted for inflation, that the federal government spent on boarding schools and assimilation from 1871 to 1969.

The broader pattern of federal funding for Native programs is long-standing. The US Commission on Civil Rights found in its 2018 *Broken Promises* report that "during the last 15 years, most federal funding for Native American programs has remained stagnant." It also found that the federal government had failed to keep accurate, consistent and comprehensive records of its own spending on these programs, and it recommended "steady, equitable, and non-discretionary funding" directly to tribal nations.

Federal language grants. The main federal source dedicated to Native languages is a set of competitive grants administered by the Administration for Native Americans, part of the US Department of Health and Human Services. They come in two broad kinds. Preservation and maintenance grants support a wide range of language projects: documentation, curriculum, teacher training, classes and community programs. Immersion grants, named for the storyteller and language advocate Esther Martinez, support language nests, immersion classrooms and similar intensive programs. A 2019 reauthorization lengthened immersion grant terms to up to five years and lowered the minimum number of participants, so that small nests could qualify.

These grants are competitive and few. In a typical year, the whole country shares a few dozen new awards or fewer. Applications are demanding. They are also among the best fits for language work, and they are worth pursuing if your office has a clear, scoped project.

As of 2026: The Administration for Native Americans' fiscal year 2025 notices offered $12 million for 15 new Preservation and Maintenance awards and $3 million for 4 new Esther Martinez Immersion awards. The fiscal year 2026 notices, posted in July 2026, offered $10 million for about 12 Preservation and Maintenance awards and $4.5 million for about 6 Esther Martinez awards, each between $100,000 and $900,000. The overall Native American Programs total was $63.5 million in fiscal year 2026, up from $60.5 million. The 2019 reauthorization authorized the Esther Martinez programs through fiscal year 2024; as of October 2026, Congress had not enacted a new reauthorization, though appropriations continued to fund the grants.

Historic preservation formula grants. Tribal Historic Preservation Offices receive a formula grant through the National Park Service's Historic Preservation Fund. Unlike competitive grants, formula grants go to every eligible office according to a set method. The THPO grant requires no match, which is a real advantage. It carries standard federal reporting requirements: a federal financial report (the SF-425), performance reporting, federal subaward reporting where it applies, and an annual accomplishments report.

The problem is dilution. The number of THPOs has grown faster than the funding line, so the average award has fallen even as offices' workloads have grown. THPOs together cover a land base exceeding 50 million acres in 30 states, and the National Association of Tribal Historic Preservation Officers has reported a significant increase in consultation work without matching staff funding. An award of around $100,000 a year funds perhaps one or two staff for a full program of review, survey and consultation.

As of 2026: THPOs grew from 211 in fiscal year 2023 to 235 eligible for fiscal year 2026, while the THPO line moved only from $23 million to $23.75 million. The average award fell from about $108,000 in 2023 to $104,000 for 221 offices in fiscal year 2024 and $100,877 for 228 offices in fiscal year 2025. The fiscal year 2026 notice distributes the enacted $23.75 million across 235 awards, with a floor of $83,000 and a ceiling of $181,000, splitting 80 percent equally and 20 percent by land area. The overall Historic Preservation Fund was $181 million to $205 million in fiscal year 2026, depending on whether congressionally directed projects are counted, and only a small share goes to THPOs.

Digital access and broadband money. In recent years, the largest pots of federal money available to tribes have been for broadband and digital access. These include infrastructure grants to build networks and "use and adoption" grants for devices, digital skills training, public computer centers and related work. Some of this money can support cultural and language programs indirectly: devices for elders, training, community technology centers, connections for remote classes. These programs tend to be large, one-time and complex, with significant administrative requirements.

As of 2026: The National Telecommunications and Information Administration (NTIA) released two related notices on June 17, 2026. The Native Entities Grant Program (NEGP) made no less than $250 million available, with expected awards of $500,000 to $2.5 million, a match of at least 10 percent (with a waiver petition allowed), a 15 percent de minimis indirect rate for applicants without a negotiated rate, and administrative and evaluation costs each capped at 10 percent. Tribal governments with a resolution approving broadband deployment under the federal BEAD program on their lands were considered first and weighted more heavily. Funds were to be spent within four years, with awards beginning in spring 2027 on a rolling basis. Its original deadline of September 17, 2026 was extended to November 17, 2026 by a July 29, 2026 technical amendment that left the other terms unchanged. That deadline will have passed for most readers of this book. The notice limits beneficiaries to members of a Native Entity but sets no residence boundary, so tribes planning to serve at-large citizens should confirm eligibility with NTIA; devices may go only to beneficiaries who complete a training program. Round 3 of the Tribal Broadband Connectivity Program, about $540 million, was released at the same time with the same extended deadline. Through November 2025, the Tribal Broadband Connectivity Program had awarded $2.24 billion to 275 projects; some previously recommended Round 2 awards were reportedly rescinded in 2025. The $42.45 billion BEAD program carries no tribal set-aside. The Digital Equity Act's competitive grant program was cancelled in May 2025; after a July 2026 court ruling, the government agreed to reinstate it and was targeting a December 2026 opening (NTIA published draft application forms for comment in August 2026; no notice had been issued as of late October 2026). Indian tribes, Alaska Native entities and Native Hawaiian organizations are eligible entities under the statute. The State Capacity Grant remained cancelled.

Philanthropy as bridge money. Private foundations and corporate giving programs fund some tribal cultural work, but at a scale far below what most people assume. According to Native Americans in Philanthropy and Candid, from 2002 to 2016 foundations gave an average of 0.4 percent of their grant dollars explicitly to Native Americans, and only about 20 percent of large foundations gave to Native causes at all. First Nations Development Institute's *Growing Inequity* (2018) found that among the top 1,000 foundations, giving to Native causes fell from $119 million in 2006 to $84 million in 2014, a fall of about 40 percent after inflation. Only about half reached Native-controlled organizations, and arts and culture received about one in seven dollars. A 2026 Brookings analysis of 2024 data found 0.88 percent of US foundation and corporate giving went to Native-serving causes or organizations, and 0.51 percent to exclusively Native-serving ones; the median grant to exclusively Native-serving causes was $20,000.

Philanthropy works best as flexible bridge money: funding a pilot that will later attract public money, covering a gap between grants, paying for something public grants will not, or supporting an institutional partnership such as the archive work described in Chapter 8. It is rarely a substitute for recurring public or tribal funding. Relationships matter more in philanthropy than in federal grants. Many foundations fund by invitation, and a conversation with a program officer before applying is often essential.

Tribal funding. The strongest base for any cultural program is the Nation's own budget, because it is not subject to a funder's priorities or a grant cycle. I return to this in Chapter 14. For now, note that a modest, recurring line in the tribal budget, even one that covers only a coordinator's salary, changes everything else: it provides match for federal grants, continuity between awards, and a signal to funders that the Nation is committed.

Reading a notice of funding opportunity. Federal grants are announced in a document called a notice of funding opportunity, or NOFO. NOFOs are long, and their most important information is often scattered. When I read one, I look for the following, in this order. Eligibility: is your Nation or office eligible? Are there restrictions on who can apply (for example, only tribal governments, not tribal nonprofits)? Deadline, and any earlier deadlines for letters of intent or registrations. Award size and number: how much, and how many awards? This tells you your odds. Match or cost share: is a match required, how much, and can it be in-kind? Allowable and unallowable costs: can you pay elders, buy equipment, pay staff? Reporting requirements: how often, and what kind? Evaluation criteria: how will applications be scored? This tells you what to emphasize. Required attachments: council resolution, budget forms, letters of support, indirect cost agreement. Contacts: the program officer's name. Call them. Most are glad to answer questions before the deadline.

The go/no-go screen. Every grant application costs time, and a small office cannot apply for everything. Before you commit, run each opportunity through a quick screen. Answer yes or no to each question.

Does it fit our charter and current priorities? Are we clearly eligible? Is the award large enough to justify the application and reporting effort? Can we meet the match, if any? Can we meet the deadline, including registrations and council approval? Do we have the staff to carry out the project if we win? Can we manage the reporting? Do we have, or can we get, the letters and attachments required? Are our odds reasonable, given the number of awards?

If you answer "no" to the first question, stop. If you answer "no" to more than two of the rest, think hard before proceeding. A "no-go" decision is not a failure. It is time saved for an opportunity that fits.

Registrations first. Federal grants require registrations that take time, sometimes weeks. If they are not current when you find a grant you want, you may miss the deadline no matter how strong your application. The key registrations include the System for Award Management (SAM.gov), which issues your Nation's Unique Entity Identifier and must be renewed every year; Grants.gov, the federal portal for most applications, which requires your SAM registration to be active and someone designated to submit; and agency-specific systems, since some agencies use their own systems for applications, payments or reporting.

Keep a "registrations first" checklist showing each registration, its expiration date, the person responsible, and a renewal reminder at least two months before expiration. Allow at least a month, and often more, to register or fix a lapsed registration.

Council resolution timing. Many grants require a council resolution approving the application. Councils meet on set schedules, often monthly, and agendas close days or weeks before each meeting. Find out your council's schedule and agenda deadlines, and build them into your plan. Working backwards from a grant deadline, you may need to have the resolution drafted six weeks ahead. Keep a template resolution ready (Appendix A).

The annual funding calendar. Pull all of this together into a funding calendar that looks a year ahead. List each known opportunity with its expected release month, deadline, award size, match, and your go/no-go decision. Add registration renewals, council meeting dates and report deadlines for current grants. Review the calendar monthly. Most federal programs that recur do so on roughly the same schedule each year, so after a year or two your calendar will become quite reliable.

The boilerplate library. Much of every grant application repeats: the Nation's history, the need, the office's capacity, its data sovereignty commitments. Write these sections once, carefully, and keep them in a boilerplate library: a folder of short, approved documents you can adapt for each application. I suggest these pieces to start: a history of the Nation and the program; a statement of need, with figures and sources from your language numbers sheet (Chapter 1); a capacity statement (staff, systems, past grants managed, partners); a data sovereignty statement (your Nation's ownership, consent and access rules); short staff biographies; and a description of your evaluation approach (Chapter 13). Review the library once a year and update the figures.

As of 2026: Federal appropriations for tribal programs were contested in 2025 and 2026. The fiscal year 2026 Interior appropriations law, enacted January 23, 2026, funded the Bureau of Indian Affairs at $2,490.3 million against a request of $1,773.7 million. Congress rejected nearly $1 billion in proposed fiscal year 2026 cuts to tribal programs. The fiscal year 2027 budget request, released April 3, 2026, again proposed deep cuts, including about 36.7 percent for the Bureau of Indian Affairs. Check the current status of appropriations before you plan.

Start Monday: 1. Check your Nation's SAM.gov registration expiration date and put a renewal reminder on your calendar two months before it. 2. Start an annual funding calendar with every grant your office currently holds or has applied for in the past two years. 3. Write or update one boilerplate section, the statement of need, with current figures and their sources.

© 2026 CRUNOD LLC. All rights reserved.

Next chapter

11. Managing a Grant: Match, Reporting and Audit

The day an award letter arrives is a good day. Celebrate it. Then, within the week, start managing it, because the work that decides whether your office can win the next grant begins now.

This is the chapter where my process-improvement background speaks most loudly, so let me state its message up front. Most small offices that struggle with grants do not struggle because the money is unavailable. They struggle because the mechanics of receiving, spending, documenting and reporting on the money consume the staff time the grant was meant to fund. The fix is not heroic effort. It is three simple disciplines: one calendar, one evidence folder, and one person accountable for each report.

Why mechanics keep small offices out. The Government Accountability Office, in a December 2024 summary of its prior work on tribal access to federal assistance (GAO-25-107674), named four systemic barriers: limited capacity, financial constraints, limited agency communication and remoteness. Within those categories, it identified several mechanics that hit small offices hardest.

Cost-share requirements, or match, block tribes that do not have matching funds available. Reimbursement-based grants, which pay only after the grantee has spent its own money, require cash up front that some tribes do not have. Small tribes may lack staff to apply at all, especially for competitive awards, and win only a small share when they do compete.

Add to these a fact every grants manager knows: reporting requirements are roughly fixed per grant, while award sizes vary. A $50,000 grant may require nearly as many reports as a $500,000 one. For an office running one formula grant and a few small competitive awards, administration therefore eats a large share of capacity.

The post-award kickoff. Within two weeks of an award, hold a kickoff meeting, even if it is just you and your finance office. Work through a checklist. Read the award document and the terms and conditions, start to finish. Note anything that differs from the application. Confirm the period of performance (start and end dates) and any budget period breaks. Confirm the approved budget, by category. Identify the match requirement, if any, and how it will be met. Identify every report, its due date and its format, and put each one on the reporting calendar. Name one accountable person for each report. Set up the grant in your finance system with its own account codes. Set up the evidence folder. Confirm how and when you will draw down or request funds, and whether the grant is advance or reimbursement. Note any special conditions, such as prior approval for certain costs or required trainings. Confirm the program officer's and grants officer's contacts. And brief everyone who will work on the grant on what it requires.

Keep the completed checklist in the evidence folder. When the auditor or the next program officer asks, you will have it.

The reporting calendar. One calendar for every grant the office holds. Not one per grant, and not on separate staff members' personal calendars. One shared calendar, with every report, every drawdown, every budget review and every deadline, color-coded by grant if you like. For each entry, show the report name, the grant, the due date, the accountable person, and a reminder date two to three weeks before. Review the calendar at every staff meeting. When a staff member leaves, the calendar stays.

Typical reports include financial reports, often the federal SF-425, which shows money received, spent and remaining; performance or progress reports describing what the project has accomplished against its plan; subaward and executive compensation reporting under the Federal Funding Accountability and Transparency Act (FFATA), where it applies; program-specific reports, such as the annual accomplishments report required of Tribal Historic Preservation Offices; and final reports at closeout.

The evidence folder. Every grant gets one evidence folder, in a shared location, with the same structure every time. I recommend these subfolders. Award: the award letter, terms and conditions, approved budget, any amendments. Application: the final submitted application. Correspondence: emails and letters with the funder, saved as files. Financial: budget-to-actual sheets, drawdown records, invoices, receipts, payroll records for grant-funded staff. Match: match and in-kind documentation. Reports: every report submitted, with confirmation of submission. Deliverables: evidence of work done, such as recordings logs, attendance records, photos of events, curriculum, publications. Approvals: prior approvals requested and received. Closeout: closeout documents.

Name files consistently, with dates. If a document is not in the evidence folder, assume it does not exist. When the office is audited or a staff member leaves, the folder speaks for the grant.

Documenting match and in-kind contributions. Many grants require a match: a share of project costs paid by the grantee or by third parties. Match can usually be cash (money from the tribal budget or another non-federal source) or in-kind (the value of donated goods, services, space or volunteer time). Federal rules on what counts as match, and how to value it, are set out in the federal cost principles (2 CFR 200.306) and in each grant's terms.

In-kind match is often where small offices have the most to offer and the least documentation. Here is how to handle the common kinds.

Volunteer hours: log every volunteer's hours, by date, task and grant, on a timesheet the volunteer signs. Value the time at a rate consistent with what the same work would be paid in your area. Keep the basis for the rate on file.

Donated space: if the Nation or another party provides space for the project without charge, document the space, the hours or months used, and a reasonable rental value, with the basis for that value.

Donated equipment or supplies: document the item, its fair market value, and the donor.

Elder honoraria: be careful here. Honoraria paid to elders from grant funds are project costs, not match. Honoraria paid from tribal funds may count as cash match if the grant allows it. Either way, document each payment: who, when, how much, for what session, and a signed receipt where appropriate. Check whether your Nation's finance office requires tax reporting for honoraria above a certain amount. Treat elders' time with respect in the paperwork as in person: make the receipt simple, and never let paperwork delay a payment.

Keep a match and in-kind log for each grant: one line per contribution, with date, type, description, value, basis for value, and reference to supporting documents in the evidence folder. Update it monthly, not at the end of the grant. Reconstructing a year of volunteer hours from memory the week before a report is due is miserable and unreliable.

The monthly budget-to-actual sheet. Once a month, compare what you have spent against what you planned, by budget category. A simple sheet shows each category; the approved budget; spending this month; spending to date; percentage of budget spent; percentage of time elapsed in the grant period; and a note on any variance.

The comparison between "percentage of budget spent" and "percentage of time elapsed" is the most useful number on the page. If you are halfway through the grant and have spent 15 percent of the budget, you are underspending and may need to adjust, or ask the funder about a no-cost extension. If you have spent 80 percent, you are on track to run out early. Either way, you want to know in month six, not month eleven.

Indirect costs. Indirect costs, sometimes called overhead or facilities and administrative costs, are the real costs of running an organization that cannot be tied to one project: accounting, human resources, office space, utilities, insurance. Many federal grants allow grantees to recover some indirect costs, either through a negotiated indirect cost rate agreement with a federal agency or, for grantees without one, through a de minimis rate.

The 2024 revision of the federal Uniform Guidance raised the de minimis rate to up to 15 percent of modified total direct costs (2 CFR 200.414(f)). For a small office, claiming indirect costs where allowed is not greedy. It recovers money the Nation is genuinely spending to support the project. Check with your finance office whether the Nation has a negotiated rate, and if not, whether the de minimis rate applies.

Some grants cap administrative or indirect costs. Read the terms carefully.

The Single Audit. Any non-federal entity that spends at least a threshold amount of federal awards in a fiscal year must have a Single Audit, an annual audit of its financial statements and federal programs. The 2024 Uniform Guidance revision, effective October 1, 2024, raised the threshold from $750,000 to $1,000,000 (2 CFR 200.501). The same revision raised the threshold for treating a purchase as equipment from $5,000 to $10,000.

The Single Audit is usually handled by the Nation's finance office, not by a cultural program. But your program's records feed into it. Clean evidence folders, current budget-to-actual sheets and documented match make the audit faster, cheaper and less stressful for everyone, and they reduce the risk of findings that could affect the Nation's future grants.

Grant terms that affect planning. Two features of recent federal programs show how terms can shape a project. Some programs cap administrative and evaluation costs separately, for example at 10 percent each, which means you must plan and document those categories carefully from the start. And some programs now allow longer grant terms than they once did. The 2019 reauthorization of the Esther Martinez immersion programs, for example, allowed grant terms of up to five years. A longer term reduces the burden of reapplying and gives intensive programs time to show results.

Prior approvals and changes. Projects change. An elder becomes ill, a partner withdraws, a cost turns out higher than expected. Most grants allow some flexibility, but significant changes in scope, key personnel, budget categories or project dates usually require the funder's prior approval. When in doubt, ask the program officer early, in writing, and file the answer in the evidence folder. Funders are almost always more forgiving of a change requested in advance than one discovered in a report.

Closeout. At the end of the grant, a closeout process confirms that the work is complete, the money is accounted for, and the reports are filed. Start planning closeout three months before the end date. A closeout checklist covers: all deliverables completed and documented; final financial report prepared and reconciled with the finance office; final performance report prepared; all match documented; any unspent funds identified and handled as the terms require; equipment purchased with grant funds inventoried and its disposition decided under the terms; any required final products, such as publications, recordings or curriculum, submitted; evidence folder complete and archived; and lessons learned written down for the next application.

That last item is not required by any funder, but it is the most valuable. Ask: what worked, what did not, what would we do differently? Put the answers in the boilerplate library for the next application.

One person accountable. I end where I began. For every report, every deadline and every grant, there should be one named person accountable. Not a department. Not "the team." One person, whose name is on the reporting calendar, who knows that the report is theirs and who will raise the alarm if they cannot meet the deadline. In a solo office, that person is you, and the discipline still helps: it reminds you, and your director, exactly how many reports rest on one pair of shoulders.

As of 2026: The Uniform Guidance thresholds cited in this chapter took effect for federal awards on or after October 1, 2024. Federal grant rules change periodically. Confirm current thresholds in 2 CFR 200 and in each award's terms.

Start Monday: 1. Set up one shared reporting calendar and enter every report due on every grant your office holds, with one accountable person named for each. 2. Create an evidence folder with the standard subfolders for your largest current grant and move existing documents into it. 3. Fill in a budget-to-actual sheet for that grant, comparing percentage spent with percentage of time elapsed.

© 2026 CRUNOD LLC. All rights reserved.

Next chapter

12. Staffing and Continuity

Every program I have ever managed had a single point of failure, and it was almost always a person. The one engineer who understood the old system. The one coordinator who knew which vendor to call. The one director who remembered why a decision had been made. When that person left, the program did not stop at once. It slowed, then drifted, then quietly lost the knowledge that had made it work.

Tribal cultural offices are especially vulnerable, because so many of them are run by one or two people, and because the knowledge those people hold is not only procedural. It includes relationships with elders, an understanding of family sensitivities, a sense of which stories belong to whom. This chapter is about keeping the program alive when a person leaves, and about caring for the people who stay.

The staffing picture. There is no national data on salaries, vacancies or turnover in tribal culture and language offices. No national survey of tribal government workforce in this field could be found. What exists are sector signals.

The best of these come from the Association of Tribal Archives, Libraries, and Museums. Its 2024 survey results found that 60 percent of tribal museums wanted more staff. Its preliminary 2026 findings, which ATALM notes are not representative, describe staffing in the field as "fragile, often solo" and identify unstable funding as the root cause of staff instability and program cuts. One rural Alaska respondent put the problem in a sentence: "There is not always the staff or funding to take advantage." Training opportunities exist; the people and money to use them often do not.

Meanwhile, the federal counterparts that tribal offices rely on have shrunk. According to the Government Accountability Office (GAO-26-108673), staff in federal Indian Affairs fell from 7,470 to 6,624 between January and July 2025, a net loss of 846 staff, or 11 percent. The Office of the Assistant Secretary for Indian Affairs lost 27 percent of its staff, and the Bureau of Indian Affairs 13 percent. Tribal leaders told GAO that service delivery was impaired. When federal offices shrink, work that they once did, or helped with, falls back on tribal offices that were already stretched.

Knowledge cannot live in one head. In the novel *The Ledger That Floats*, when the man who kept a veterans' post's history dies, most of it goes with him, leaving a water-stained ledger and a box of unlabeled photographs. It is fiction, but most readers who have worked in small organizations will recognize it. The goal of this chapter is to make sure that when someone leaves your office, for any reason, the next person finds a desk manual and a labeled archive rather than a ledger and a box.

The desk manual. A desk manual is a written guide to doing a job. It is not a policy manual or a job description. It is the practical, how-we-actually-do-it guide that a new person would need to keep the work going if they arrived tomorrow and you were unreachable.

A good desk manual for a cultural program covers the following. Purpose and charter: what the program is for and a copy of the current charter. Who's who: key people inside and outside the Nation, with roles and how to reach them, including elders and families, with appropriate sensitivity. Recurring tasks: what happens daily, weekly, monthly, quarterly and annually, with step-by-step instructions for each. Systems: where files are stored, how backups work, how to use each tool, where the equipment is. Grants: a summary of each current grant, its reporting calendar and its evidence folder location. Protocols: the consent forms, sensitivity review workflow, access tiers and restricted-content list. Relationships: notes on key relationships, such as which elders prefer phone calls, which families need to be consulted, which partners need careful handling. Keep this section respectful and confidential. History: a short account of the program's past, major decisions and why they were made. And open matters: anything in progress.

Start writing the desk manual in your first month, as Chapter 2 suggested, and add to it every time you do something for the first time. Aim for a working draft in six months. Review it once a year. A desk manual is never finished, and that is fine.

Cross-training. Even in a two-person office, each person should be able to do the other's essential tasks, at least at a basic level. A cross-training matrix makes this visible. List the essential tasks down the side and the staff (and key volunteers) across the top. In each box, mark whether the person can do the task independently, can do it with the manual, or cannot do it. Any task that only one person can do is a risk. Plan to train a second person, and schedule a time for them to do it under supervision.

In a solo office, cross-training means identifying someone outside the office, perhaps a colleague in another department, a trusted volunteer or a staff member in a partner Nation's office, who could keep the most critical tasks going in an emergency. At minimum, someone else should know where the desk manual is, how to reach the elders on the priority list, and how to file an urgent grant report.

Passwords and credentials. The most common way a departing staff member takes a program's knowledge with them is through passwords. The office's cloud storage, the email account used for a grant system, the social media accounts, the archive login, the registration in a federal system: if only one person knows these, the program is one resignation away from being locked out of its own materials.

A credential handover procedure should do several things. Keep a secure, shared record of every account the office uses, with the account name, purpose and who has access. A password manager designed for organizations, with shared vaults and access controls, is the safest approach; a sealed envelope in a locked safe is a low-tech alternative. Never store passwords in an ordinary document or spreadsheet. Use accounts that belong to the office, not to individuals. Register program accounts with an office email address, not a personal one. Give at least two people access to every critical account. When someone leaves, change the passwords to every account they had access to, on their last day, and remove their access. Review the record every six months.

This is unglamorous work, and many offices put it off. Please do not. I have seen more programs lose access to their own records through a forgotten password than through any other single cause.

Volunteers and interns. Volunteers and interns can extend a small office's reach enormously. They can scan, transcribe, catalog, organize events, help elders with devices and support classes. They also need supervision, training and clear boundaries, especially when they handle sensitive material.

Write a short role description for each volunteer role, covering what the volunteer will do; what they will not do (for example, no recording without staff present, no access to restricted material); the time commitment; the training required; who supervises them; and what they will gain (experience, references, a stipend if available). Pair each role description with the volunteer agreement from Chapter 8. Students from tribal colleges and universities, young citizens home for the summer, and retired citizens with time to give can all be excellent volunteers. Many of them become future staff.

Growing teachers and staff from within. The most durable programs grow their own people. Hawaiian language leaders, as noted in Chapter 5, have made the point that teachers cannot be imported: parents must learn the language and become the teachers. The same is true of staff. A learner who sat with elders during recording sessions, an apprentice who finished a master-apprentice program, a volunteer who spent a summer cataloging photographs: these are the people most likely to understand the work and stay with it.

Make the path visible. Show learners and volunteers how they could move into paid roles. Support them in training, certifications and education. Include succession in your planning: for every key role, identify someone who could step into it in two or three years.

The succession checklist. When a staff member is leaving, or might leave, work through a succession checklist. Is the desk manual current? Have all credentials been documented and shared? Have elders and families been told, respectfully, about the change and introduced to the person who will continue the work? Have partners and funders been notified, where required? Are all grant reports current, and is the reporting calendar up to date? Are all files in shared locations, not on personal devices? Has the departing person walked their successor or an interim person through each recurring task? Has a debrief been held to capture anything not written down? Have the departing person's accounts and access been closed or transferred?

If possible, overlap the departing and arriving staff for at least two weeks. If not, have the departing person record short walk-through explanations of key tasks, and leave contact information if they are willing to answer questions.

The new-staff 30-day plan. When someone new joins, a 30-day onboarding plan helps them become effective quickly and builds their confidence. In week one, they read the charter and desk manual, meet the director and key colleagues, and get access to systems. In week two, they shadow existing work, review grant files and the reporting calendar, and learn the consent and review process. In week three, they meet key elders and families, with an introduction from someone the elders trust, and meet key partners. In week four, they take on recurring tasks under supervision, review the 90-day plan, and identify questions and gaps in the desk manual.

The introduction to elders deserves special care. Elders have often formed a relationship with the previous staff member. A new face can be unsettling. An introduction in person, by someone the elder knows and trusts, makes all the difference.

Caring for staff. Cultural work carries an emotional weight that other office jobs do not. Staff who record elders form relationships with people near the end of their lives. When elders pass, staff grieve, often while also handling the practical consequences: the restricted recordings, the family's requests, the gap in the program. Staff who work with historical materials encounter painful histories of loss and injustice. Staff in small offices carry a sense of responsibility for their Nation's future that can be heavy.

I am not a counselor, and I will not pretend to give clinical advice. But a few practices help. Acknowledge loss openly. When an elder passes, give staff time to grieve and to take part in their Nation's customs. Do not expect business as usual the next day. Build in support. Know what employee assistance or counseling services the Nation offers, and make sure staff know too. Some offices bring in a respected elder or spiritual adviser after a loss, according to the Nation's customs. Share the load. No one person should carry every relationship and every difficult conversation. Rotate where possible. Set limits. A solo coordinator who is on call every evening and weekend will not last. Agree on reasonable hours and protect them. Celebrate. Mark the good days: a first recording session, a learner's first conversation, a completed grant. Small celebrations sustain people through hard seasons.

As of 2026: The GAO report on 2025 workforce reductions in federal Indian Affairs was published in January 2026. Federal staffing may have changed further since then. Check current GAO and agency reports.

Start Monday: 1. Create a secure record of every account your office uses, and confirm that at least two people can access each critical one. 2. Start, or update, your desk manual by writing step-by-step instructions for the one recurring task only you know how to do. 3. Fill in a cross-training matrix for your office's ten most essential tasks and identify the biggest single-person risk.

© 2026 CRUNOD LLC. All rights reserved.

Next chapter

Part Four · 13. Measuring What Matters

At the end of a long year, a director once asked me to help her write the annual report for her language program. I sat down with her and her records. She had attendance sheets, sign-in lists, receipts and a box of photographs. She knew the program had done good work. But when she tried to answer the question her council would ask, "Are more people speaking our language than a year ago?", she could not say. The program had counted everything except the thing that mattered most.

That is the most common measurement problem I see, and it is not the office's fault. Funders ask for enrollment numbers, attendance and hours delivered, because those are easy to count and easy to compare. Councils and communities want to know something harder: is the language coming back? Are the elders' stories being kept? Are citizens far from home reconnecting? This chapter is about measuring those things in ways a small office can manage.

The evidence gap. I want to begin honestly. Across the field, most programs count enrollment, not speakers produced. First Nations Development Institute's Native Language Immersion Initiative supported 32 programs from 2018 to 2021, serving about 6,000 students, but its evaluation summary reported no aggregate proficiency data across those programs. No tribal language app has published data on retention or proficiency. The federal 10-Year Plan sets a goal of increasing proficient speakers by 25 percent before 2036, but there is no national baseline to measure that against, because no national speaker database exists.

This is not a criticism of those programs. Measuring language proficiency is hard, takes time and requires trained assessors. Small offices rarely have the capacity. But it does mean that measurement itself is part of the help offices need. An office that can show, with simple and credible evidence, that its program produces speakers is in a far stronger position than one that can show only attendance.

What to count. I suggest thinking about measurement in three layers, from easiest to most meaningful.

Activity: what the program did. Sessions held, hours delivered, recordings made, items digitized, newsletters sent. These are easy to count and necessary for grant reports.

Reach: who the program served. People enrolled, attendance, elders recorded, citizens away from home reached. Also fairly easy, and important for showing who benefits.

Results: what changed. Learners whose proficiency improved. Recordings reviewed and usable, not only hours captured. Citizens away from home who took part in a program for the first time. Families who received copies of an elder's recordings. Teachers trained who are now teaching. These are harder to count, and they matter most.

A good measurement system includes all three layers, but puts its effort into results. For each program area, choose one or two result measures you can collect reliably, and collect them consistently.

The one-page logic model. A logic model is a simple diagram that connects what a program puts in to what it hopes to achieve. Funders often ask for one, and even when they do not, writing one clarifies your thinking. It has five columns.

Inputs: what you put in. Staff time, money, equipment, elders' knowledge, partners. Activities: what you do. Recording sessions, classes, digitization, newsletters. Outputs: the direct products. Number of sessions, recordings, learners, items, issues. Outcomes: the changes you expect in the short and medium term. Learners' proficiency improves; recordings are available to families and teachers; citizens away from home connect with the program. Impact: the long-term change you are working toward. More of the Nation's culture spoken, taught, told and remembered by more of its own people.

Keep it to one page. Write it in plain language. Show it to elders and staff, not just funders, and ask whether it describes the program they recognize.

An indicator menu. Here is a menu of indicators by program area. Choose a few from each area your office works in. Do not try to use them all.

Elder recording: elders recorded this year, and as a percentage of the priority list; recording hours captured; recordings reviewed and approved by speaker or family, as a percentage of those captured; recordings with complete metadata; sessions with a learner present; and families who received copies.

Language classes and intensive programs: learners enrolled, and active after three and six months; contact hours per learner, especially in intensive programs; learners whose proficiency rose by at least one level on the program's scale; apprentices who completed the program; and former learners now teaching or mentoring.

Digitization and collections: items digitized, by type and priority; items with complete metadata and assigned access tier; at-risk items stabilized; photographs identified through elder sessions; and materials received through digital return.

Governance: outside holders covered by a written agreement, as a percentage of all known holders; recordings with signed consent forms; and annual data audit and export test completed.

Citizens away from home: at-large citizens on the contact list; first-time participants from outside tribal lands; and newsletter open rates, where your email tool reports them.

Continuity: desk manual updated this year; critical tasks that at least two people can perform; and accounts with at least two authorized users.

Measuring proficiency simply. Proficiency is the hardest result to measure and the most important for language programs. You do not need a formal testing program to begin. A simple approach works for many offices.

Choose a scale. The proficiency guidelines of the American Council on the Teaching of Foreign Languages (ACTFL) are widely used and give a common language for funders; the Cherokee Nation's master-apprentice program, for example, has set a target of at least intermediate-mid on the ACTFL scale. Some Nations have developed their own scales, which may fit their language and values better. A simple five-point scale, from "knows a few words" to "can hold a full conversation on everyday topics," is a reasonable start.

Assess regularly. Every three to six months, have a fluent speaker hold a short conversation with each learner, perhaps ten to fifteen minutes, and rate them on the scale. Use the same speaker where possible, for consistency. Record the result in the learner progress record from Chapter 5.

Report change, not just level. The most persuasive result is movement: "Of our twelve apprentices, nine moved up at least one level this year." That tells a council and a funder far more than "we had twelve apprentices."

Be honest about limits. A self-built scale is not a standardized test. Say so in reports. Funders respect honesty about methods more than inflated claims.

The quarterly tracking sheet. Collect your chosen indicators on a simple quarterly tracking sheet: one row per indicator, one column per quarter, with a target for the year. Fill it in at the end of each quarter from your logs, rosters and catalog. It should take an hour or two if your records are kept as described in earlier chapters. The quarterly sheet does three things. It makes annual reporting easy, because the numbers are already collected. It shows trends early, so you can adjust. And it builds a record over years, which is how a program proves it works.

Listening to elders and learners. Numbers do not tell the whole story. A short feedback form, offered to elders and learners once or twice a year, adds the voices that numbers miss. For elders, ask: How have you felt about the recording sessions? Is there anything we should do differently? Is there knowledge you would like to share that we have not asked about? Would you like copies of your recordings for your family? For learners, ask: What has helped you learn most? What has made it harder? How often do you use the language outside the program? What would help you keep going? Offer the form on paper and online, and offer to fill it in with elders in conversation. Read every response. Quote them, with permission, in reports. A single sentence from an elder about why the program matters can carry more weight with a council than a page of figures.

Reporting results. Report to three audiences, in three different ways.

To council: a short written summary each quarter, with the key indicators and one or two stories. An annual briefing in person. Councils are busy. Lead with results and keep it brief.

To the community: an annual community report, in plain language, with photographs (where consent allows), stories and a few key numbers. Distribute it in every channel from Chapter 9, so that elders at home and citizens away from home both see it. Thank the elders, learners, volunteers and families by name where they wish to be named.

To funders: the reports required by each grant, filed on time. Use your quarterly tracking sheet to fill them in. Include results, not just activities, even if the funder's form does not ask for them.

An annual community report template, in Appendix A, gives a structure: a message from the director or an elder; what we did this year; what changed; stories; thanks; what comes next; and how to get involved.

Returning to the measure of success. In the Author's Note, I offered my own measure of success for this work: more of each Nation's culture spoken, taught, told and remembered by more of its own people, and understood and respected by more people beyond it, shared on the Nation's terms, for generations to come.

That sentence can serve as the impact column of your logic model. Each of its parts suggests indicators. "Spoken" points to proficiency and speakers. "Taught" points to teachers and classes. "Told" points to recordings and storytelling. "Remembered" points to archives, metadata and access. "By more of its own people" points to reach, including citizens away from home. "On the Nation's terms" points to governance: consent, agreements, access tiers. You do not need to measure everything. But if each part of that sentence has at least one indicator behind it, your program is measuring what matters.

Start Monday: 1. Draft a one-page logic model for your main program, with all five columns. 2. Choose three indicators from the menu, including at least one result measure, and set up a quarterly tracking sheet for them. 3. Write down one story from this quarter, with permission, that shows what the program has changed for a learner, an elder or a family.

© 2026 CRUNOD LLC. All rights reserved.

Next chapter

14. Sustainability: After the Grant Ends

The best time to plan for the end of a grant is the day you apply for it. The second-best time is today.

I have watched many programs, in many fields, follow the same arc. A grant arrives. Staff are hired, equipment is bought, a system is built. For three years the program thrives. Then the grant ends, and nothing replaces it. Staff leave. The system goes unmaintained. Within two years, little remains but a final report and some files nobody can open. The program did not fail because its work was poor. It failed because nobody planned for year four.

This chapter is about planning for year four on day one.

Total cost of ownership. Every tool, system, archive and program has costs that continue after the initial purchase or grant. Program managers call the full picture the total cost of ownership. For a small cultural office, it usually includes subscriptions and licenses, often annual; storage, which grows every year as you add recordings and scans; backup and the 3-2-1 copies (Chapter 6); maintenance and updates for any system you run; technical support, whether from a vendor, a partner or a consultant; equipment replacement, since phones, microphones, computers and drives wear out every few years; staff time to run, maintain and use the system; training for new staff and volunteers; and format migration, because over decades files must be moved to new formats and new storage as old ones become obsolete.

Before you adopt any tool or launch any program, fill in a total-cost-of-ownership worksheet: list each cost, estimate it for each of the next five years, and add them up. Then ask the hard question: where will the money for years four and five come from? If there is no answer, either find one or choose something cheaper.

Free tools deserve special scrutiny. A tool that costs nothing to start may charge later, may change its terms, may disappear, or may be paid for in ways that are not money, such as rights to use your content. Apply the vendor-neutral scorecard from Chapter 8 to every free tool.

Why community and platform sustainability are inseparable. The developers of the open-source archive platform discussed in Chapter 8 reached a conclusion I want to repeat here, because it is the heart of sustainability. They found that the standard open-source model, in which users maintain the software themselves, failed for tribal users who lacked "the resources, infrastructure, and programming skills." They concluded that community sustainability and platform sustainability are "inseparable." A system survives only if the community using it survives as a community of users, and a community's digital archive survives only if someone keeps the system running.

Their answer was the hybrid model described earlier: an institution maintains the core, regional hubs provide training, and each tribe sets its own needs. That model shows up again and again in programs that last. It is the third and fourth parts of the five-part model from the Introduction: an institution that carries the technical maintenance, and regional peer training.

For your office, the practical question is: who will carry the technical maintenance after the grant? If the answer is "me, in my spare time," the program is at risk. If the answer is "a university partner, under a written agreement that keeps ownership with the Nation," or "a regional consortium of tribal offices sharing a technician," or "a vendor whose cost is in the tribal budget," the program has a fighting chance.

Tribal self-funding as the strongest base. The most sustainable funding for a cultural program is the Nation's own. Tribal funding is not subject to a federal agency's priorities, a foundation's strategy change or a grant cycle. It signals to funders that the Nation is committed, it provides match for federal grants, and it carries the program through gaps between awards.

The Cherokee Nation offers the best-known example. Its Durbin Feeling Language Preservation Act, made permanent in January 2024, sets a minimum of $18 million a year for language work; the Council raised the 2024 budget above $20 million, and the Nation reports $182 million invested under the Act to date. Very few Nations can commit at that scale, and I do not suggest that they should try to match it. The lesson is the principle, not the amount: a dedicated, recurring line in the tribal budget, protected by law or resolution, gives a program a floor that grants cannot.

For a small Nation, that floor might be the salary of one coordinator, or the annual cost of storage and backup, or the honoraria for elders. Even a modest line changes the program's position. When you brief your council, present the total cost of ownership and show which costs a tribal line would cover. Councils are more willing to fund a specific, recurring, modest cost than a vague request for "support."

Building a base of support. Beyond tribal and grant funding, a program can build a base of support among citizens, families and allies. This is not mainly about money, though small donations can help. It is about people who care whether the program survives and will speak up for it. A base of support might include citizens away from home who take online classes and value them; families whose elders were recorded and who received copies; former learners and apprentices; volunteers; partner institutions; and local businesses and organizations.

Keep these people informed, through the channels in Chapter 9 and the annual community report from Chapter 13. Invite them to events. Ask them to tell the council what the program means to them. When budget decisions are made, a program with a visible base of support is far harder to cut.

In the novel *The Veteran's Quiet Bridge*, one phone call and a supper in a community-hall basement grow into a network of neighbors who keep each other's stories. It is fiction, about a very different community, but it illustrates something true about sustainability: it is built from people, not only from funding. A program that has woven itself into the lives of families and citizens has a kind of resilience that no grant can buy.

Diversified funding. No single source should carry the whole program. A diversified funding map shows each funding source, what it pays for, how reliable it is, and when it ends. Sources typically include the tribal budget (most reliable, often smallest); formula grants, such as historic preservation funds (fairly reliable, small); competitive federal grants (unreliable, larger, time-limited); philanthropy (unreliable, flexible); partner contributions (in-kind support from universities or institutions); and earned income, where appropriate and consistent with the Nation's values, such as fees for workshops or sales of publications.

Map the program's costs against these sources. Make sure the essential, recurring costs, such as core staff, storage and backup, rest on the most reliable sources. Use competitive grants and philanthropy for projects, pilots and expansions, not for the core. When a time-limited grant pays for something that must continue, plan from day one which reliable source will take it over.

The annual export and restore drill. Once a year, run a full export and restore drill. It is the single best test of whether your Nation can always leave with everything. Choose a date and put it on the calendar every year. Export a meaningful sample of your materials, or all of them if feasible, from every system that holds them, including metadata, consent records and access settings. Restore the export to a different location, such as a separate computer or storage account. Check that the files open in standard software, that metadata is complete and readable, and that access settings and consent records survived. Time it and note the cost. Write a short report of what worked and what did not, and fix what did not.

The drill tests your backups, your partners' compliance with their export obligations, and your own ability to move. If a vendor raises its prices, a partner closes or a system stops working, the drill means you already know how to leave.

Survivorship bias, one more time. I mentioned in Chapter 8 that no United States tribal app or archive has published a public account of shutting down when its grant ended. It is worth repeating here, because it shapes how you should read every success story. When you hear about a thriving program, you are hearing from a survivor. You rarely hear from the programs that quietly stopped. Learn from the survivors, but assume that the failures were real and numerous, and plan as if your program could become one of them unless you act.

The sustainability plan. Bring all of this together in a short sustainability plan, ideally written before the first grant is submitted and revised every year. It should cover: the program's purpose and core activities; the total cost of ownership, for five years; current funding sources and what each covers; the costs that must continue after current grants end, and the plan for each; who carries the technical maintenance, and under what agreement; the base of support, and how it is maintained; the results of the most recent export and restore drill; and risks and how they will be managed. Share it with your director and council. Update it every year. When you apply for a grant, include a summary. Funders increasingly ask how a project will continue after their money ends, and a written plan is the best possible answer.

As of 2026: The Cherokee Nation's language investment figures cited here come from the Nation's January 2024 announcement and its September 2026 State of the Nation address. Check the Nation's current releases before citing them.

Start Monday: 1. Fill in a total-cost-of-ownership worksheet for the main system your office uses to keep recordings, for the next five years. 2. Draw a diversified funding map showing each funding source, what it pays for and when it ends, and circle any essential cost that rests on a time-limited grant. 3. Put the date of your first annual export and restore drill on the calendar, and name who will run it.

© 2026 CRUNOD LLC. All rights reserved.

Next chapter

Closing: Spoken, Taught, Told and Remembered

I began this book with a belief, and I want to end with it. Records can hold a culture for a while, but they cannot bring it back once it has gone quiet. Only living people, speaking and teaching, can carry it forward.

Everything in these pages has been in service of that belief. The charters and decision charts, the consent forms and access tiers, the backup rules and reporting calendars, the desk manuals and export drills: none of them is the point. They are the sturdy, unglamorous frame that lets the real work happen and keeps it from vanishing when a person leaves, a grant ends or a server fails. The real work is an elder speaking and a learner listening. It is a teacher in a classroom, a family hearing a grandmother's voice, a young citizen in a distant city taking part in the Nation's language for the first time.

When I look across the evidence gathered for this book, three priorities stand out. They are not the only things that matter, but they are the things that existing programs, funders and partners most often leave undone, and they are where a small office can make the greatest difference.

The first is to record elders and pair them with learners now. The window is measured in years. An average speaker near 75, a life expectancy that is still years shorter than the national figure, and no national count of what has already been lost: those facts do not allow for waiting until conditions are perfect. Begin with a phone, a consent form and a learner. Record generously, review carefully, and remember that the goal is speakers and relationships, not only files. The Yuchi story shows what a few months at an elder's side can do.

The second is to lower the administrative cost of every dollar. Grant mechanics, not grant availability, keep many small offices out. Match, reimbursement, fixed reporting costs and complex registrations eat the staff time that grants were meant to fund. One calendar, one evidence folder, one accountable person for each report; a desk manual and a cross-trained colleague; a boilerplate library and a go/no-go screen: these turn a solo office from one that drowns in its awards into one that can use them. Every hour saved on paperwork is an hour that can go to an elder's kitchen table.

The third is to put ownership, consent, limits on automated systems and export rights in writing before anyone outside the Nation touches a recording. A tribe's jurisdiction stops at its borders, and data does not. The written agreement is the main protection a Nation has. Tribal ownership; no secondary use, including training any computer system, without written tribal consent; tribe-set access tiers; provenance; full export in open formats; verified deletion on exit. Six terms, on one page, before the first file leaves the office.

Behind those three priorities sits the five-part model from the Introduction: tribal ownership in writing, tribe-set access rules, an institution that carries the technical maintenance, regional peer training, and an intensive human program at the center, with technology as the on-ramp. No single office builds all five parts at once. But every office can build the first two in its first year, and can look for partners and peers to help with the third and fourth. The fifth, the human program, is already there wherever an elder and a learner sit together.

I also want to say something about the two audiences. A Nation's citizens are no longer, for the most part, gathered in one place. Most live away from home. They are connected, but often disengaged, and they ask for timely information in the channels they already use. The elders at home are engaged, but often poorly connected, and they need someone to sit beside them. An office that serves only one of these groups misses most of its people. An office that serves both, with the same information at the same time in every channel, holds the Nation together across distance. That is not a technical achievement. It is an act of care.

Finally, I want to return to the image that has stayed with me through all the reading and listening behind this book. It comes from the documented failure in Chapter 7, but I want to turn it around.

Imagine a family, some years from now, coming to your office. Their grandmother has passed. They remember that she sat with your program once, in her kitchen, and told stories in the language. They ask whether they can hear her voice.

In one version of that moment, the answer is that the recording belongs to someone else, or that nobody knows where it is, or that the file will not open, or that the person who made it left years ago and took the password with them. That is the version this book has tried to help you prevent.

In the other version, you open your catalog. You find the recording, with her name and the date, the learner who sat with her, the consent she signed and the review she approved. You check its access tier and see that her family may hear it. You make them a copy. And you tell them, because it is true, that her voice belongs to them, and to their people, and that it will be kept on the Nation's terms for as long as the Nation chooses.

That is the moment all of this work is for. More of each Nation's culture spoken, taught, told and remembered by more of its own people, and understood and respected by more people beyond it, shared on the Nation's terms, for generations to come.

Thank you for doing this work. Start Monday.

© 2026 CRUNOD LLC. All rights reserved.

Next chapter

Appendix A: Templates and Checklists

These templates collect the tools described in each chapter. Copy them, adapt them and put any that touch consent, ownership or access through your own council and legal review. Replace English terms with your own language's terms wherever that fits.

A1. What We Hold Today inventory (Chapter 1)

Known speakers and knowledge-holders (confidential):
- Name or speaker code
- Approximate age
- Knowledge held (language, songs, places, histories, crafts, other)
- Health and willingness, as known
- Date of last contact
- Learner currently paired, if any

Existing recordings:
- Description
- Format (cassette, reel, video tape, digital)
- Approximate date
- Speaker, if known
- Current holder (office, family, institution, individual)
- Written agreement in place? (yes, no, unknown)

Collections:
- Type (photographs, documents, maps, objects, records)
- Approximate quantity
- Location
- Condition (stable, monitor, act soon, act now)

Classes and programs:
- Name of activity
- Who runs it
- How often it meets
- Number attending
- Funding source

Gaps and risks:
- What worries you most, in a sentence each

A2. Language numbers sheet (Chapter 1)

For each figure your office uses:
- The figure
- What it counts (languages, speakers, home speakers, fluent speakers, learners)
- Source
- Year of the data
- Scope (Nation, state, United States, North America)
- Threshold (first-language, conversational, any proficiency)
- Date you last checked it

Five questions before using any language statistic: Who counted? What did they count? What is the scope? What year? What threshold?

A3. One-page program charter (Chapter 2)

- Program name
- Purpose (one or two sentences)
- Scope: included in this phase
- Scope: not included in this phase
- Accountable owner
- Decision-makers (see decision chart)
- Timeline: start date, three or four milestones, end of phase
- Resources: staff time, budget, equipment, space
- Measures of success (two or three, countable)
- Risks and responses
- Approved by, title, date

A4. Decision chart (Chapter 2)

Down the left, list decisions. Across the top, list roles. Mark each box R (responsible), A (accountable, exactly one per decision), C (consulted before) or I (informed after).

Suggested decisions:
- Approve program charter
- Choose elders to approach first
- Approve consent forms and protocols
- Approve a recording for use beyond the office
- Assign access tier
- Sign agreement with an outside partner
- Approve a grant application
- Approve public communications

Suggested roles: coordinator, director, cultural committee or elders' council, tribal council, speaker's family, legal counsel.

A5. 90-day start plan (Chapter 2)

Days 1 to 30, set up:
- Charter approved
- Consent form and protocol drafts in review
- Recording kit assembled and tested
- Storage and backup set up
- Elder priority list drafted
- Desk manual started

Days 31 to 60, first sessions:
- Three to five sessions with learners present
- Each recording reviewed with speaker or family
- Process adjusted from experience

Days 61 to 90, rhythm and review:
- Regular schedule in place
- Report to director, committee and elders
- Charter revised if needed
- Next 90 days planned

A6. One-pilot selection worksheet (Chapter 2)

For each candidate pilot, score 1 to 5:
- Urgency (will waiting a year cause permanent loss?)
- Feasibility (can we do it with what we have now?)
- Visibility (will finishing it show the program works?)
- Foundation (will it build systems later projects need?)
Total the scores. Highest total is the pilot. On a tie, choose the higher foundation score.

A7. Consent form elements (Chapter 3; see Appendix B for sample language)

- Speaker name and preferred contact
- Family, clan or community sign-off, where protocol requires
- Purpose of the recording
- Uses allowed
- Uses excluded, including use to train any computer system without written tribal consent
- Review step before any use beyond storage
- Access tier (completed after review)
- How to withdraw or narrow consent, and whom to contact
- Signatures and date
- Staff member and learner present

A8. Restricted-content flag list (Chapter 3)

For each sensitive or restricted category your Nation identifies:
- Category (for example: songs, ceremonies, healing knowledge, family histories, sacred places, genealogy)
- Restriction type (season, gender, age, standing, family, mourning, other)
- Who may hear or see it
- Who decides exceptions
- How items are marked in the catalog

A9. Sensitivity review workflow (Chapter 3)

1. Record under signed consent; hold at most restricted level.
2. Log with metadata; flag known sensitive content.
3. Speaker or family review; record edits and restrictions requested.
4. If flagged, cultural reviewer or committee review.
5. Assign access tier; log reviewer and date.
6. Release at assigned tier only.
7. Revisit at least yearly and after any death or major change.

A10. Protocol policy outline (Chapter 3)

- Purpose and authority
- Definitions
- Sensitive and restricted categories
- Consent requirements (individual, family, community)
- Review and release process
- Access tiers
- Withdrawal of consent
- Requests from partners and researchers
- Storage and security
- Periodic review of the protocol

A11. Elder priority list (Chapter 4; confidential)

For each elder:
- Speaker code
- Age
- Health considerations (as shared by elder or family)
- Unique knowledge
- Willingness
- Priority rank
- Paired learner
- Next contact date

A12. Recording kit list (Chapter 4)

- Dedicated smartphone with ample storage
- External microphone (clip-on or compact directional)
- Small tripod or phone stand
- Battery pack and cables
- Headphones
- Printed consent forms, metadata sheets, pens
- Storage device for same-day transfer
- Gift, food or honorarium as appropriate

A13. Home-visit checklist (Chapter 4)

- Called ahead and confirmed the day before
- Asked whether a family member should be present
- Brought what is customary
- Arrived on time; allowed time for conversation
- Explained consent clearly; answered questions; offered time to decide
- Kept session to a comfortable length; watched for fatigue
- Thanked the elder; explained next steps and review
- Left contact information
- Followed up as promised

A14. Session metadata sheet (Chapter 4)

- Date and place
- Speaker code(s)
- Others present, including learner
- Language(s) spoken
- Topics (a few words)
- Sensitive content flagged (yes, no; category)
- Consent form reference
- Equipment used
- File names
- Questions to follow up
- Completed by, date

A15. File-naming convention (Chapter 4)

Pattern: YYYY-MM-DD_SPEAKERCODE_SESSION_TYPE
Example: 2026-03-14_SPK017_S02_audio
Rules: no spaces, no special characters, speaker code rather than name, rename on the day of transfer.

A16. Learner in the room pairing plan (Chapter 4)

For each elder:
- Learner(s) assigned
- Family, dialect or gender considerations
- Minimum number of sessions committed
- Learner tasks (equipment, questions, transcription, summary)
- Check-in date

A17. Hours-to-fluency worksheet (Chapter 5)

For each format (weekly class, weekend camp, summer intensive, master-apprentice, other):
- Hours per week
- Weeks per year
- Hours per year per learner
- Years to 2,000 hours (2,000 divided by hours per year)
- Years to 2,880 hours

A18. Master-apprentice pairing agreement outline (Chapter 5)

- Speaker, apprentice and program contact
- Hours per week, months, target total hours
- Language-only rule and exceptions
- Activities planned
- Payment for speaker and apprentice
- Hours and progress records
- Office support provided
- Pausing or ending the pairing
- Recording consent
- Signatures and date

A19. Class roster, attendance and progress records (Chapter 5)

Roster: learner name, contact, start date, program.
Attendance log: date, session, learner, minutes attended.
Progress record: learner, assessment date, assessor, level on program scale, notes.
Session record: date, teacher, content covered, materials used.

A20. Remote-class readiness checklist (Chapter 5)

- Teacher and learners have a tested connection
- Session length suited to online attention (shorter than in person)
- Audio-only option for weak connections
- Recording consent settled for the session
- Access limited to approved learners
- Materials sent ahead in small files
- Someone available to help an elder join

A21. Digitization priority matrix (Chapter 6)

For each group of material, score 1 to 5:
- Risk of loss or damage
- Value to the Nation
- Expected use
Add or multiply. Any "act now" condition goes first regardless of score.

A22. Condition triage sheet (Chapter 6)

- Type of material
- Quantity
- Storage location and conditions
- Visible damage
- Urgency: stable, monitor, act soon, act now
- Specialist needed? (yes, no)

A23. Minimum metadata fields (Chapter 6)

- Unique identifier
- Title or short description
- Date (approximate allowed)
- People shown or speaking (if appropriate)
- Place
- Creator
- Source (holder of original)
- Rights and ownership
- Access tier
- Labels applied
- Catalogued by, date

A24. Recommended file formats (Chapter 6)

- Photographs and documents, master: TIFF, 400 to 600 ppi for photographs, 300 to 400 ppi for documents
- Photographs and documents, access: JPEG or PDF
- Audio, master: WAV (24-bit, 96 kHz for analog transfers)
- Audio, access: MP3 or similar
- Video: high-quality MP4 at original resolution, original tapes kept where possible
- Born-digital text for keeping: PDF/A
Confirm choices with a partner archive.

A25. 3-2-1 backup plan (Chapter 6)

- What is backed up
- Copy 1: location, how often, by whom
- Copy 2: different kind of storage, location, how often, by whom
- Copy 3: different physical place, how often, by whom
- Quarterly restore test: date, files tested, result

A26. Outside-holder inquiry letter (Chapter 6)

Dear [name or department],

I write on behalf of the [office name] of the [Nation name], which holds authority under [law, resolution or charter] for the care of our Nation's cultural materials. We are seeking to identify recordings, photographs, documents and other materials relating to our Nation, our people, our language or our lands that may be held in your collections.

Could you tell us whether you hold such materials, and if so, share a list with any existing descriptions? We would also be grateful to learn about your policies on access, copies and community control. We refer you to the Protocols for Native American Archival Materials, endorsed by the Society of American Archivists in 2018, as a framework for this conversation.

We would welcome a conversation at your convenience. Please contact [name, title, phone, email].

With thanks,
[Name, title]

A27. Access tier matrix (Chapter 7)

Tiers down the side: public, citizens, family, elders or knowledge-holders, restricted (add your own).
Actions across the top: view, listen, download, copy, share, publish.
Mark each box allowed, allowed with permission, or not allowed.

A28. Data governance policy outline (Chapter 7)

- Purpose and legal authority
- Definitions
- Ownership
- Collection and approval
- Consent
- Access tiers
- Storage and security
- Sharing with outside parties and required terms
- Prohibited uses, including use to train any computer system without written tribal consent
- Retention and deletion
- Roles and responsibilities
- Complaints and enforcement
- Review of the policy

A29. Council resolution checklist (Chapters 7 and 10)

- Legal counsel review completed
- Cultural committee or elders consulted
- No conflict with existing tribal law or policy
- Resolution language states clearly what is approved
- Responsible person named
- Cost stated
- Review date set
- Agenda deadline for the target council meeting confirmed

A30. Written-terms checklist (Chapter 7)

- Tribal ownership of all materials, copies and derived works
- No secondary use without written tribal consent, including training any computer system
- Tribe-set access tiers applied and enforced
- Provenance metadata kept and never stripped
- Full export in open formats at any time and at exit
- Verified deletion on exit, certified in writing

A31. Annual data audit (Chapter 7)

- Every copy of important materials located
- Every outside holder covered by a current agreement
- Access tiers reviewed
- Departed staff and volunteer access removed
- Backups restored and checked
- Export test completed
- Summary reported to director and council

A32. Vendor-neutral evaluation scorecard (Chapter 8)

Score each 0 (not met) to 3 (fully met), with one sentence of evidence:
- Tribal ownership in writing
- Access controls that enforce your tiers
- Full export in open formats
- Consent records stored with items
- Phone and low-bandwidth use
- Support and training
- Cost after the grant
- Exit plan and certified deletion
Reject for anything above the public tier if ownership, export or exit scores 0.

A33. Academic partner MOU outline (Chapter 8)

- Parties, signed through the tribal government
- Purpose and scope
- Ownership with the Nation
- The six written terms (A30)
- Publication and tribal review of drafts
- Credit for the Nation and speakers
- Rules for students and student-created material
- Storage and security
- Accountable contact on each side, and notice when it changes
- Reporting
- Term, renewal, termination
- Dispute resolution

A34. Volunteer agreement and role description (Chapters 8 and 12)

Agreement:
- Role and tasks
- Confidentiality
- Ownership of anything created belongs to the Nation
- No copies, no posting, no personal use, no use to train any computer system
- Training required
- Supervisor
- Ending the role: return of materials and removal of access
- Signatures and date

Role description:
- What the volunteer will do
- What the volunteer will not do
- Time commitment
- Training
- Supervisor
- What the volunteer gains

A35. Exit and export test (Chapter 8)

- System tested
- Sample exported (content, metadata, consent, access settings)
- Opened on a separate computer in standard software? (yes, no)
- Metadata complete? (yes, no)
- Time taken and cost
- Problems found and actions

A36. Two-audience channel plan (Chapter 9)

Citizens away from home: website or portal, monthly email newsletter, online classes, official social media, email or text reminders.
Elders at home: printed newsletter mailed early, phone calls, local radio, in-person classes and gatherings, tech-help sessions.
Rule: same information, same day, every channel.

A37. At-large citizen contact and consent form (Chapter 9)

- Name
- Citizen identification, where appropriate
- Mailing address, email, phone
- Preferred channels
- Interests
- Consent to be contacted, and how to unsubscribe
- Date

A38. Low-bandwidth design checklist (Chapter 9)

- Works on an older phone
- Pages load on a slow connection
- Audio playable without video
- Offline download where the tier allows
- Large, clear text and easy buttons
- Simple sign-in
- Tested by an elder while you watched

A39. Elder tech-help session plan (Chapter 9)

- Regular time and place
- One helper for two or three elders
- One task per session
- Large-print take-home steps
- Phone number for help between sessions

A40. Content calendar (Chapter 9)

For each channel: what goes out, on which day, who prepares it, who approves it.

A41. Annual funding calendar (Chapter 10)

For each opportunity: funder, program type, expected release month, deadline, award size, match, go/no-go decision, accountable person. Add registration renewals, council meeting dates and report deadlines.

A42. Go/no-go grant screen (Chapter 10)

Yes or no:
- Fits our charter and priorities
- Clearly eligible
- Award worth the effort
- Match can be met
- Deadline can be met, including registrations and council approval
- Staff to carry out the project
- Reporting can be managed
- Attachments obtainable
- Reasonable odds
Stop if the first answer is no. Reconsider if more than two others are no.

A43. Registrations first checklist (Chapter 10)

For each registration (SAM.gov, Grants.gov, agency systems):
- Account holder
- Expiration date
- Renewal reminder (two months ahead)
- Person responsible
Allow at least a month for new or lapsed registrations.

A44. Boilerplate library contents (Chapter 10)

- Nation and program history
- Statement of need with sourced figures
- Capacity statement
- Data sovereignty statement
- Staff biographies
- Evaluation approach
- Lessons learned from past grants
Review yearly.

A45. Post-award kickoff checklist (Chapter 11)

- Award and terms read in full
- Differences from application noted
- Period of performance confirmed
- Approved budget confirmed
- Match requirement and plan confirmed
- Every report on the reporting calendar with one accountable person
- Account codes set up
- Evidence folder set up
- Advance or reimbursement confirmed
- Special conditions noted
- Program and grants officer contacts saved
- Team briefed

A46. Reporting calendar entry (Chapter 11)

- Report name
- Grant
- Due date
- Accountable person
- Reminder date (two to three weeks before)
- Submitted date and confirmation

A47. Evidence folder structure (Chapter 11)

Award; Application; Correspondence; Financial; Match; Reports; Deliverables; Approvals; Closeout.

A48. Match and in-kind log (Chapter 11)

- Date
- Type (cash, volunteer time, space, equipment, supplies)
- Description
- Value
- Basis for value
- Supporting document reference

A49. Monthly budget-to-actual sheet (Chapter 11)

For each budget category: approved budget, spent this month, spent to date, percentage spent, percentage of time elapsed, variance note.

A50. Closeout checklist (Chapter 11)

- Deliverables complete and documented
- Final financial report reconciled
- Final performance report
- Match documented
- Unspent funds handled per terms
- Equipment inventoried and disposition decided
- Required products submitted
- Evidence folder archived
- Lessons learned written

A51. Desk manual template (Chapter 12)

- Purpose and charter
- Who's who
- Recurring tasks with steps (daily, weekly, monthly, quarterly, annual)
- Systems, storage and backups
- Grants summary
- Protocols
- Relationship notes (confidential)
- Program history and key decisions
- Open matters

A52. Cross-training matrix (Chapter 12)

Tasks down the side; people across the top. Mark: independent, with manual, or not yet. Any task with only one "independent" is a risk to address.

A53. Credential handover procedure (Chapter 12)

- Secure record of every account (name, purpose, who has access)
- Organizational password manager or sealed record in a locked safe
- Office email for every program account
- Two authorized users for every critical account
- Passwords changed and access removed on a departing person's last day
- Review every six months

A54. Succession checklist (Chapter 12)

- Desk manual current
- Credentials documented and shared
- Elders and families informed and introduced to successor
- Partners and funders notified where required
- Reports current
- Files in shared locations
- Walk-through of recurring tasks
- Debrief held
- Accounts closed or transferred

A55. New-staff 30-day onboarding plan (Chapter 12)

Week 1: charter, desk manual, introductions, system access.
Week 2: shadowing, grant files, consent and review process.
Week 3: meet elders and families with a trusted introduction; meet partners.
Week 4: supervised recurring tasks; review 90-day plan; list questions.

A56. Staff wellbeing note (Chapter 12)

- Acknowledge loss; allow time for customs
- Know and share available support services
- Share difficult relationships and conversations
- Agree on reasonable hours
- Celebrate milestones

A57. One-page logic model (Chapter 13)

Inputs | Activities | Outputs | Outcomes | Impact

A58. Quarterly tracking sheet (Chapter 13)

One row per indicator; columns for annual target, Q1, Q2, Q3, Q4, year total, notes.

A59. Elder and learner feedback form (Chapter 13)

Elders: How have the sessions felt? What should change? What knowledge have we not asked about? Would your family like copies?
Learners: What helps most? What makes it harder? How often do you use the language outside the program? What would help you keep going?

A60. Annual community report template (Chapter 13)

- Message from the director or an elder
- What we did this year
- What changed
- Stories (with permission)
- Thanks
- What comes next
- How to get involved

A61. Total-cost-of-ownership worksheet (Chapter 14)

For each cost, estimate years 1 to 5: subscriptions, storage, backup, maintenance, support, equipment replacement, staff time, training, format migration. Total each year. Name the source of funds for years 4 and 5.

A62. Diversified funding map (Chapter 14)

For each source: what it pays for, reliability (high, medium, low), end date. Circle any essential recurring cost resting on a time-limited source.

A63. Annual export and restore drill (Chapter 14)

- Date
- Systems exported
- Restored to
- Files open in standard software? (yes, no)
- Metadata, consent and access settings intact? (yes, no)
- Time and cost
- Problems and fixes

A64. Sustainability plan outline (Chapter 14)

- Purpose and core activities
- Five-year total cost of ownership
- Current funding sources and coverage
- Costs that continue after grants, and the plan for each
- Technical maintenance and agreement
- Base of support
- Latest drill results
- Risks and responses

© 2026 CRUNOD LLC. All rights reserved.

Next chapter

Appendix B: Sample Policy Language

Please read this first. The sample language in this appendix is not legal advice. It is a starting point offered by a program manager, not a lawyer, and it has not been reviewed for any particular Nation's laws, constitution or customs. Before you adopt any of it, adapt it with your own legal counsel and present it to your own governing body, whether that is a council, a cultural committee or another authority under your Nation's law. Where your Nation's law, protocols or customs differ from anything here, they govern. Replace bracketed text with your own details, and consider using your own language's terms for key concepts.

B1. Consent for recording

I, [speaker name], agree to be recorded by the [office name] of the [Nation name] on [date] at [place].

Purpose. I understand that this recording is being made to [purpose, for example: keep my knowledge of our language and stories for our people and future generations].

Uses allowed. I agree that this recording may be used for the following purposes only: [check those that apply: storage in the Nation's archive; access by citizens of the Nation; use in the Nation's language classes; copies for my family; other].

Uses not allowed. I do not agree to the following uses unless I, or after my passing my family, and the Nation give new written consent: public posting on the internet; commercial use of any kind; use by outside researchers; and use to train any computer system or automated tool, or sharing with anyone who would do so.

Others with authority. I have been asked whether any of the knowledge I share belongs to my family, clan or community. [Check one: It does not. / It does, and the following person has also been asked and agrees: name and signature.]

Review. Before this recording is used for anything beyond storage, I or a family member I name will be given the chance to review it and ask for changes or restrictions. I name [family member] to review it if I am not able to.

Access. After review, the office will assign an access level under the Nation's protocol and will tell me which level was assigned.

Withdrawal. I may withdraw or narrow this consent at any time by contacting [name, title, phone]. If I do, the office will restrict or remove the recording as I ask, from its archive and from any copies it controls. I understand that a recording already played at a public event cannot be unheard, but it will not be used again.

Signatures. Speaker, date. Family or community sign-off, where required, date. Staff member, date. Learner present, if any.

B2. Access tiers

Section 1. All cultural materials held by or for the Nation shall be assigned one of the following access tiers after review under the Nation's protocol.

(a) Public: may be shared with any person, including in publications and online, with attribution to the Nation and, where the speaker or family wishes, to the speaker.

(b) Citizens: may be shared only with enrolled citizens of the Nation [and any other persons the Nation designates, such as descendants or spouses].

(c) Family: may be shared only with the family of the speaker or subject, as identified by the family.

(d) Elders and knowledge-holders: may be shared only with persons designated under the Nation's protocol.

(e) Restricted: may be shared only with the specific permission of [designated authority], including materials restricted by season, gender, ceremonial standing or mourning.

Section 2. Until reviewed, all materials shall be held at the Restricted tier.

Section 3. Access tiers shall be reviewed at least once a year and after the death of a speaker or subject, and may be changed only by [designated authority].

Section 4. No system may be used to store or share materials above the Public tier unless it can enforce these tiers.

B3. Secondary use, including automated systems

Section 1. No person, organization or partner may use any cultural material, recording, transcript, translation, image or data of the Nation for any purpose other than the purpose for which it was provided, without the prior written consent of the Nation given through [designated authority].

Section 2. Without limiting Section 1, no person, organization or partner may use any such material to train, test or improve any computer system or automated tool that learns from data, or provide any such material to any other party for that purpose, without the prior written consent of the Nation.

Section 3. Every agreement with an outside party that involves the Nation's materials shall include the terms of this section.

Section 4. A breach of this section shall entitle the Nation to [remedies, for example: immediate termination of the agreement, return and verified deletion of all materials, and other remedies available under the agreement and applicable law].

B4. Ownership, export and deletion (for agreements with outside parties)

Ownership. The Nation owns all materials and data provided to or created under this agreement, including all copies, transcripts, translations, metadata and derived works. [Partner] holds them only as a steward for the purposes stated in this agreement and acquires no ownership, license or other right except as expressly stated.

Access tiers. [Partner] shall apply and enforce the access tiers assigned by the Nation to each item and shall not change them.

Provenance. [Partner] shall keep all provenance and descriptive metadata provided by the Nation and shall not remove or alter it.

Export. At any time on [number] days' written notice, and at the end of this agreement, [Partner] shall provide to the Nation complete copies of all materials and metadata in open, standard formats specified by the Nation, at [no cost / a cost not exceeding amount].

Deletion. Within [number] days after the end of this agreement, or on the Nation's written request, [Partner] shall permanently destroy all copies of the materials and data in its possession or control, including backups, and shall certify the deletion in writing, signed by an authorized officer.

Survival. These terms survive the end of this agreement.

B5. Withdrawal of consent (office procedure)

When a speaker or family withdraws or narrows consent, the office shall, within [number] days: record the request and the date; change the access tier of the affected materials as requested; remove the materials from any shared or public location the office controls; notify any partner holding copies and require compliance under its agreement; confirm the actions in writing to the speaker or family; and record the actions taken in the catalog.

© 2026 CRUNOD LLC. All rights reserved.

Next chapter

Glossary

Access tier. A level of permission assigned to an item that determines who may see, hear, copy or share it, such as public, citizens, family, elders or restricted.

ACTFL. The American Council on the Teaching of Foreign Languages, whose proficiency guidelines (novice, intermediate, advanced, superior, with sublevels) are widely used to describe language ability.

At-large citizen. An enrolled citizen of a Nation who lives away from its lands.

CARE Principles. Principles for Indigenous data governance: Collective Benefit, Authority to Control, Responsibility and Ethics. Published in 2020.

Cost share or match. The portion of a project's costs that a grantee must pay from non-grant sources, in cash or in kind.

De minimis indirect rate. A standard indirect cost rate, currently up to 15 percent of modified total direct costs under federal rules, that a grantee without a negotiated rate may charge.

Digital return. Making high-quality digital copies of materials held by outside institutions and returning them to the community, usually with the community gaining control over their description and access.

Esther Martinez. The name given to federal immersion grants for Native languages, supporting language nests, immersion classrooms and similar programs.

FAIR principles. Principles that data should be Findable, Accessible, Interoperable and Reusable. The CARE Principles were written to complement them.

FFATA. The Federal Funding Accountability and Transparency Act, which requires reporting on certain subawards and executive compensation.

Immersion. Teaching in which the language is the medium of instruction for most or all activities, rather than a subject taught in another language.

In-kind. A non-cash contribution to a project, such as volunteer time, donated space or donated equipment, valued at a reasonable rate.

Indigenous data sovereignty. The right of a nation to govern the collection, ownership and application of its own data.

Language nest. An immersion setting for very young children, usually from infancy to about age five, staffed by fluent speakers.

Local Contexts Labels. Traditional Knowledge and Biocultural Labels applied by communities, through the Local Contexts Hub, to signal their own rules for the use of digital materials.

Master-apprentice. A program pairing a fluent speaker with a committed adult learner for many hours a week, using only the language.

Metadata. Information about an item, such as its date, creator, subject, place, rights and access tier.

MOU. Memorandum of understanding: a written agreement setting out the terms of a partnership.

NOFO. Notice of funding opportunity: the document announcing a federal grant program and its rules.

Provenance. The history of an item: who made it, under what agreement, who has held it and what has been done to it.

SF-425. The standard federal financial report used by many federal grant programs.

Single Audit. An annual audit required of non-federal entities that spend at least a threshold amount of federal awards in a year, currently $1,000,000.

THPO. Tribal Historic Preservation Officer or Office, which assumes historic preservation responsibilities on tribal lands under federal law.

Traditional Knowledge Labels. One kind of Local Contexts Label, used to convey a community's expectations for attribution, access and use of traditional knowledge.

3-2-1 backup. A rule for protecting files: at least three copies, on at least two kinds of storage, with at least one in a different physical place.

© 2026 CRUNOD LLC. All rights reserved.

Next chapter

Sources and Further Reading

Figures in this book were checked against the sources below, which were accessed in October 2026. They date from 2014 to 2026. Check each source for newer figures before quoting it.

Language counts, speakers and immersion

US Department of the Interior, US Department of Education and US Department of Health and Human Services. 10-Year National Plan on Native Language Revitalization. December 2024. https://www.bia.gov/sites/default/files/media_document/final_12032024_10-year_national_plan_on_native_language_revitalization.508.pdf

US Department of the Interior. Release announcing the 10-Year National Plan. https://www.doi.gov/pressreleases/biden-harris-administration-releases-10-year-national-plan-native-language

Voice of America. "Native American Languages Fight for Survival." October 2023. https://www.voanews.com/a/native-american-languages-fight-for-survival/7302835.html

First Nations Development Institute. Native Language Immersion Initiative webinar materials. 2025. https://www.firstnations.org/wp-content/uploads/2025/04/NLII-Campaign-2025-Webinar-powerpoint-1.pdf

First Nations Development Institute. Native Language Immersion Initiative evaluation summary for practitioners. 2022. https://www.firstnations.org/wp-content/uploads/2022/01/NLII-Eval-Summary-PRACTITIONERS.pdf

US Census Bureau. 2017-2021 American Community Survey language use tables, press release. June 2025. https://www.census.gov/newsroom/press-releases/2025/2017-2021-acs-language-use-tables.html

US Census Bureau. "Native American Language Use." June 2025. https://www.census.gov/library/stories/2025/06/native-american-language-use.html

Cherokee Nation. Citizens by State 2026. https://www.cherokee.gov/media/ltbmrjmy/cherokee-citizens-by-state-2026.pdf

Cherokee Nation, Anadisgoi. Permanent reauthorization of the Durbin Feeling Language Preservation Act. January 2024. https://www.anadisgoi.com/index.php/government-stories/chief-hoskin-signs-permanent-reauthorization-of-historic-durbin-feeling-language-preservation-act

Cherokee Nation, Anadisgoi. Annual first-language Cherokee speaker gathering. November 2024. https://anadisgoi.com/index.php/culture-stories/cherokee-nation-hosts-annual-first-language-cherokee-speaker-gathering

Cherokee Nation, Anadisgoi. Registration and photo ID field events, 2026 schedule. https://anadisgoi.com/index.php/government-stories/cherokee-nation-registration-and-photo-id-field-events-reach-18-000-in-2025-2026-schedule-announced

Cherokee Nation, Anadisgoi. 2026 State of the Nation. September 2026. https://anadisgoi.com/index.php/government-stories/2026-sotn

Cherokee Phoenix. Cherokee Nation and others on language struggles during the pandemic. 2021. https://www.cherokeephoenix.org/culture/cherokee-nation-others-weigh-in-on-language-struggles-during-pandemic/article_6237b518-bfb9-11eb-87f3-93c4578aecf2.html

Cherokee Phoenix. New Cherokee speakers succeed despite pandemic. August 2022. https://www.cherokeephoenix.org/culture/new-cherokee-speakers-succeed-despite-pandemic/article_32a44a84-1cbd-11ed-8cb9-93a4b845848a.html

Scientific American. "Can an App Save an Ancient Language?" September 2016. https://www.scientificamerican.com/article/can-an-app-save-an-ancient-language/

Rawlins, testimony before the Senate Committee on Indian Affairs. August 22, 2018. https://www.indian.senate.gov/sites/default/files/Final%20Testimony%20Rawlins.SCIA_.22.Aug_.2018.pdf

Elders and the pandemic

Administration for Community Living. 2021 Profile of American Indians and Alaska Natives Age 65 and Older. https://acl.gov/sites/default/files/Profile%20of%20OA/AIANProfileReport2021.pdf

National Center for Health Statistics. Provisional life expectancy estimates for 2021. August 2022. https://www.cdc.gov/nchs/pressroom/releases/20220831.html

NVSR 72-12 (NCHS, November 2023): life tables for the United States in 2021. https://www.cdc.gov/nchs/data/nvsr/nvsr72/nvsr72-12.pdf

NVSR 75-5 (NCHS, August 2026): life tables for the United States in 2024. https://www.cdc.gov/nchs/data/nvsr/nvsr75/nvsr75-05.pdf

ABC17 and CNN. Cherokee Nation saving vaccine doses for fluent speakers. January 2021. https://abc17news.com/news/national-world/2021/01/12/only-about-2000-people-speak-the-cherokee-language-fluently-the-tribe-is-saving-some-vaccine-doses-for-them/

Minnesota Public Radio News. COVID-19 deaths of tribal elders. April 2021. https://www.mprnews.org/story/2021/04/09/our-hearts-are-heavy-covid19-deaths-of-tribal-elders-leave-a-void

Cronkite News. Shawnee language classes move online. August 2021. https://cronkitenews.azpbs.org/2021/08/13/shawnee-language-classes-move-online-during-covid-19-pandemic/

PBS NewsHour. Native communities in Oklahoma raced to preserve culture. https://www.pbs.org/newshour/health/as-covid-swept-the-state-native-communities-in-oklahoma-raced-to-preserve-culture

Consent, protocols and data sovereignty

Native Nations Institute, University of Arizona. US Indigenous Data Sovereignty Network. 2016. https://nni.arizona.edu/news/us-indigenous-data-sovereignty-network

Carroll et al. "The CARE Principles for Indigenous Data Governance." Data Science Journal 19:43, 2020. https://datascience.codata.org/articles/10.5334/dsj-2020-043

National Congress of American Indians. Resolution KAN-18-011. https://archive.ncai.org/resources/resolutions/support-of-us-indigenous-data-sovereignty-and-inclusion-of-tribes-in-the-development-of-tribal-data

National Congress of American Indians. Comments on federal policy, March 15, 2025. Posted in the federal docket at https://files.nitrd.gov/90-fr-9088/

Protocols for Native American Archival Materials. Northern Arizona University. https://www2.nau.edu/libnap-p/protocols.html

Society of American Archivists. Council endorsement of the Protocols. August 2018. https://www2.archivists.org/statements/saa-council-endorsement-of-protocols-for-native-american-archival-materials

Havasupai Tribe v. Arizona Board of Regents, Arizona Court of Appeals, 2008. https://nill.narf.org/bulletins/state/documents/2009/havasupai_v_az_board_of_regents.html

Arizona Board of Regents and plaintiffs. Joint release on settlement. April 21, 2010. https://turtletalk.blog/wp-content/uploads/2010/04/havasupai-asu-press-release.pdf

State Press. Board of Regents settles lawsuit with Havasupai Tribe. April 2010. https://statepress.com/article/2010/04/abor-settles-lawsuit-with-havasupai-tribe-over-blood-samples

History.com. Why the Navajo Nation banned genetic research. 2017. https://www.history.com/articles/why-the-navajo-nation-banned-genetic-research

Cherokee Nation, Anadisgoi. Cherokee Nation signs first policy on automated tools. August 2025. https://anadisgoi.com/index.php/government-stories/cherokee-nation-signs-first-ai-policy-vows-to-protect-language-culture-while-exploring-work-efficiency

Context. Indigenous communities and language data. April 2023. https://www.context.news/ai/nz-us-indigenous-fear-colonisation-as-bots-learn-their-languages

ICT News. Standing Rock to sue to gain Lakota language materials. December 2024, updated November 2025. https://ictnews.org/news/standing-rock-to-sue-to-gain-lakota-language-materials/

Partnership models and archives

Plateau Peoples' Web Portal, About. https://plateauportal.libraries.wsu.edu/about

Christen, Merrill and Wynne. D-Lib Magazine, May/June 2017. https://dlib.org/dlib/may17/christen/05christen.html

Washington State University College of Arts and Sciences. January 2019. https://cas.wsu.edu/?p=28891

KOSU. Grant bolsters Miami Tribe language revitalization work at Miami University. February 2025. https://www.kosu.org/race-culture/2025-02-14/2-1-million-grant-bolsters-myaamia-centers-indigenous-language-revitalization-efforts

Ojibwe People's Dictionary, About. https://ojibwe.lib.umn.edu/about

KUNC. Online Indigenous language learning opens doors for far-flung tribal citizens. February 2021. https://www.kunc.org/community/2021-02-11/online-indigenous-language-learning-opens-doors-for-far-flung-tribal-citizens

Association of Tribal Archives, Libraries, and Museums, via Native Nations Institute. 2024 reports. https://nni.arizona.edu/news/sustaining-and-advancing-indigenous-cultures-reports-published-association-tribal-archives

Association of Tribal Archives, Libraries, and Museums. Preliminary survey findings. January 2026. https://www.atalm.org/wp-content/uploads/2026/02/Preliminary-Survey-Findings.pdf

Citizens away from home and broadband

Harvard Kennedy School, Ash Center. Harvard Project databook (third edition), September 2025. https://ash.harvard.edu/wp-content/uploads/2025/09/Databook-Third-Edition-2025-09-07-1.pdf

Urban Indian Health Institute. https://www.uihi.org/urban-indian-health/

Kopriva. "Impacts of the Relocation Program on Native American Migration and Fertility." Journal of Economic History 84(1), 2024. https://www.cambridge.org/core/services/aop-cambridge-core/content/view/91BEFB71F4213DB977C2FFBA821B41EA/S0022050724000032a.pdf/impacts-of-the-relocation-program-on-native-american-migration-and-fertility.pdf

National Council of Urban Indian Health. 2020 Census undercount. August 2023. https://ncuih.org/2023/08/28/urban-american-indian-undercount-in-the-2020-census-went-underreported/

Schultz, Rainie and Starks. Native Nations Institute. June 2015. https://nnigovernance.arizona.edu/sites/default/files/2023-03/residence%20-%20community%20engagement%20and%20citizenship.pdf

Government Accountability Office. GAO-22-104421. June 2022. https://www.gao.gov/products/gao-22-104421

Congressional Research Service. R48563. June 2025. https://www.everycrsreport.com/reports/R48563.html

Federal Communications Commission. 2026 Section 706 Report. https://docs.fcc.gov/public/attachments/FCC-26-55A1.pdf

Benton Institute for Broadband and Society. Summary of the FCC 2026 report. https://www.benton.org/blog/did-fcc-answer-broadband-deployment-question-congress-asked

Federal Reserve Bank of Minneapolis. Home internet access on tribal lands after the Affordable Connectivity Program. September 2024. https://www.minneapolisfed.org/article/2024/after-the-affordable-connectivity-program-home-internet-access-on-tribal-lands-ticks-up

Fierce Network. What happened to NTIA's tribal broadband money. November 2025. https://www.fierce-network.com/broadband/what-happened-ntias-tribal-broadband-money

Broadband Breakfast. Previously recommended tribal broadband awards rejected. November 2025. https://broadbandbreakfast.com/previously-recommended-tribal-broadband-awards-rejected.md

Funding, grants and federal rules

National Telecommunications and Information Administration. Amended NEGP notice of funding opportunity, 2026. https://files.simpler.grants.gov/opportunities/cc2a99b9-6669-4f45-b233-9189edfaa59e/attachments/5ea32550-336a-45c3-81f8-3e3a777914d0/NTIA_Amended_NEGP_NOFO_2026.pdf

National Telecommunications and Information Administration. TBCP3 and NEGP technical amendment, July 2026. https://files.simpler.grants.gov/opportunities/3c1f350b-8a60-43ae-baa3-a5a20959845f/attachments/4da59d2a-bae4-4e2c-bc3f-cde6cd1b5b3a/TBCP3_NEGP_NOFO_Technical_Amendment_2026.pdf

National Telecommunications and Information Administration. TBCP3 NOFO primer. June 2026. https://broadbandusa.ntia.gov/sites/default/files/2026-06/TBCP3_NOFO_Primer.pdf

American Library Association. Digital Equity Act FAQ. Updated September 2026. https://www.ala.org/advocacy/federal-resources/broadband-policy/digital-equity-resources/DEA-FAQ

Benton Institute. NTIA takes first public step toward reopening digital equity competitive grants. https://www.benton.org/headlines/ntia-takes-first-public-step-toward-reopening-digital-equity-competitive-grants

Administration for Native Americans. FY2025 Native language funding opportunities. https://acf.gov/ana/news/ana-announces-fy2025-grant-funding-opportunities-native-languages-and-environmental

Simpler.Grants.gov. FY2026 Preservation and Maintenance notice. https://simpler.grants.gov/opportunity/361743

Simpler.Grants.gov. FY2026 Esther Martinez Immersion notice. https://simpler.grants.gov/opportunity/361742

US Department of Health and Human Services. FY2027 Administration for Children and Families congressional justification. https://www.hhs.gov/sites/default/files/fy-2027-acfc-cj.pdf

Congress.gov. S.256, Esther Martinez Native American Languages Programs Reauthorization Act (P.L. 116-101). https://www.congress.gov/116/bills/s256/BILLS-116s256enr.htm

National Conference of State Historic Preservation Officers. Historic Preservation Fund. https://ncshpo.org/issues/historic-preservation-fund/

National Conference of State Historic Preservation Officers. SHPOs and THPOs. February 2026. https://ncshpo.org/wp-content/uploads/2026/02/SHPOs-and-THPOs.pdf

National Association of Tribal Historic Preservation Officers. Testimony to House Appropriations. May 2024. https://docs.house.gov/meetings/AP/AP06/20240508/117270/HHRG-118-AP06-Wstate-GrussingV-20240508.pdf

National Trust for Historic Preservation. FY26 Interior appropriations bill. https://savingplaces.org/historic-preservation-fund/updates/fy26-interior-appropriations-bill

Congressional Research Service. R48726. March 2026. https://www.everycrsreport.com/files/2026-03-19_R48726_0aaabda061f3a0b9f24663d5fe42647ce23dc98e.html

Hobbs Straus. General Memorandum 26-019. April 2026. https://hobbsstraus.com/wp-content/uploads/2026/04/GM-26-019.pdf

Friends Committee on National Legislation. April 2026 Native American legislative update. https://www.fcnl.org/updates/2026-04/april-2026-native-american-legislative-update

US Commission on Civil Rights. Broken Promises press release. December 2018. https://usccr.gov/press/2018/12-20-PR-Broken-Promises.pdf

Native Americans in Philanthropy. Philanthropic giving to Native Americans. November 2020. https://nativephilanthropy.org/2020/11/17/native-americans-are-2-of-u-s-population-but-receive-0-4-of-philanthropic-dollars/

First Nations Development Institute. Growing Inequity. 2018. https://www.firstnations.org/wp-content/uploads/2026/06/Growing-_Inequity_WEB.pdf

Brookings Institution. Investing in Indian Country. August 2026. https://www.brookings.edu/articles/investing-in-indian-country/

Government Accountability Office. GAO-25-107674, Tribal Issues: Barriers to Access to Federal Assistance. December 2024. https://www.gao.gov/products/GAO-25-107674

Office of Management and Budget. Guidance for Federal Financial Assistance, 89 FR, April 22, 2024 (2 CFR 200.501 and 200.414(f)). https://www.federalregister.gov/documents/2024/04/22/2024-07496/guidance-for-federal-financial-assistance

Government Accountability Office. GAO-26-108673, Indian Affairs: Information on 2025 Workforce Reductions. January 2026. https://www.gao.gov/products/gao-26-108673

© 2026 CRUNOD LLC. All rights reserved.

Next chapter

Acknowledgments

[Placeholder: thanks to colleagues, advisers and readers to be added before publication.]

I am grateful to the staff of tribal cultural, language, historic preservation, library, archive, museum and grants offices whose work and generosity shaped my understanding of this field, and to the researchers, reporters and organizations whose public work is cited in these pages. Any errors are mine.

This book has not yet been reviewed by Native readers. Tribal readers: if any page here misstates a fact, leaves out something important, or sits poorly with your Nation's customs, please write and tell me. I will fix it in the next edition.

© 2026 CRUNOD LLC. All rights reserved.

Next chapter

About the Author

Vikas Bakshi founded CRUNOD LLC, a one-person company. His career has been spent in program management, process improvement, project delivery and technology integration, helping organizations turn good intentions into programs that run, report and last. He works alongside tribal and cultural offices on the practical systems that support cultural preservation. He also founded EternalCurrent.Online.

© 2026 CRUNOD LLC. All rights reserved.

If this book stayed with you

The Last Speaker's Granddaughter
Learning to Say Kesi Right

Another ECO Library story with Native American roots.

About this book

Back cover of Spoken, Taught, Told and Remembered

The people who hold a Nation's language, songs and stories are elders, and the office trying to record, teach and keep that knowledge often has one or two staff, a patchwork of small grants and no time to spare.

This guide gives that office a program manager's toolkit, put to work for the work that matters: consent and protocols before the first recording, recording elders with a learner in the room, language classes built around an intensive core, digitizing fragile collections, written control over the Nation's data, choosing partners without losing ownership, reaching citizens who live far from home, and finding, managing and outlasting grants.

Every chapter ends with three things a one-person office can do in a single week. Sixty-four templates and checklists, sample policy language and a full list of sources make it ready to use, adapt and photocopy.

It starts from one belief: only a people owns its culture, and the job of any tool or helper is to make that ownership easier to exercise.

Members can react to this book and talk about it in the members' Library.

Ownership

© 2026 CRUNOD LLC. All rights reserved.

This book is owned by CRUNOD LLC. It is free to read on EternalCurrent.Online. No part of it may be copied, reprinted, sold, recorded or adapted, including as audio, film or translation, without written permission from CRUNOD LLC.

EternalCurrent.Online

© 2026 CRUNOD LLC. All rights reserved.

More books in the Library

Every book is reviewed against our content standards before it reaches a shelf. Want to write for the Library? Ask to become an author.