Language Keepers: A Documentary Film Process

Language Keepers: A Documentary Film Process for Stimulating Passamaquoddy-Maliseet Language Documentation and Revival

Ben Levine and Robert M. Leavitt

[PDF Download]

This paper describes a National Science Foundation Endangered Language Initiative entitled “Language Keepers: Audio Visual Documentation of Passamaquoddy-Maliseet Group Discourse Centered on Historic Places as a Data Source for Dictionary Expansion, Linguistic Analysis, and Teaching Resources.” The authors are the co-principal investigators of this project.

Begun in August 2006 on the eastern Maine Passamaquoddy reservations (Pleasant Point and Indian Township), Language Keepers employs an unusual multi-disciplinary approach to address an increasingly common problem in endangered language documentation and reclamation: PassamaquoddyMaliseet is no longer spoken in public.

There has been no formal assessment of Passamaquoddy-English language shift in the last three decades, but anecdotal reports on language use are highly consistent in suggesting rapid decline of Passamaquoddy as the language that (at least in public) life is lived in. When David Francis, co-author of the Passamaquoddy Dictionary, was asked when was the last time he had a Passamaquoddy conversation outside the home, he could not recall. “On a trip once a few years ago, we tried to have a conversation only in Passamaquoddy, and we kept going into English. English is so powerful!” Linguist Philip LeSourd of Indiana University, original developer of the Passamaquoddy dictionary, has remarked that when he began the dictionary in the 1970s everyone on the reservation spoke Passamaquoddy all the time. When he returned after an absence, he was “astonished” at how few people were willing or able to use the language. Allen Sockabasin (2007), writer and teacher and former tribal governor, notes that many people cannot continue a conversation after the opening exchanges. In spite of a school based bilingual education program there have been few if any new speakers since 1960.

While a complete examination of the reasons for language shift are not within the scope of this paper, it is important to note that the Passamaquoddy reservations have experienced a doubling of their population after the 1980 land claim settlement. Many of those who returned had lived off the reservation for 20 or more years; they spoke English almost exclusively and had little contact with traditional culture. Thus, in the span of a few decades, the language of the communities shifted.

Filming public discourse
For the purposes of our work, “public discourse” is any use of the language outside the family circle. It ranges from friends having brief conversations in the supermarket parking lot or gossiping, to a group playing cards or picking berries in a field, to thoughtful discussions about social change. Depending on the composition of the group, “public discourse” may take place in a private home. Clearly, the loss of public discourse in Passamaquoddy is a symptom of the language’s decline. It deprives Native-language teachers and learners of a necessary resource and interrupts the normal intergenerational transmission critical to language stabilization.

Documenting public discourse poses challenges different from those encountered in ethnographic documentation, interviewing, and elicitation as methods of linguistic research. In the former, there is an ongoing spontaneous production of language for the researcher to record from whatever observational distance is deemed appropriate. In the latter, it is the linguist who produces situations for the recording of one-on-one conversation. Here a high level of precise linguistic description is facilitated by questioning subjects as they speak or immediately after a conversation.

Since public discourse is not currently available to the Passamaquoddy, an investigator must convene group discourse for documentation in a way that is deemed natural by the participants and that will also yield significant data. As a filmmaking approach, we employ a “direct cinema” documentary style that is an outgrowth of cinéma vérité. This means that we do not try to achieve a significant social distance between the camera and the subject to avoid influencing the interaction. It is just the opposite. We are in overt collaboration with the group from the beginning. Some participants have seen previous footage and know that the cameraman will physically be part of the group, shooting in an intimate style to capture not only the language but also the nuances of body language and physical interactions, as well as the activity in which speakers are engaged. This is a widely accepted documentary style, and even among camera-shy people, it is not uncommon to hear, “I completely forgot you were filming.” In addition, with wireless, high sensitivity microphones, we are able to capture an accurate linguistic record along with images that convey a vivid sense of how people live everyday life in the language.

The Passamaquoddy dictionary
Prior to this project, language documentation had been an ongoing activity on the two Passamaquoddy reservations since the language first began its precipitous decline, in the 1960s.

The Passamaquoddy-Maliseet dictionary project began in the late 1970s, when linguist Philip LeSourd began compiling a dictionary database for the Wabnaki Bilingual Education Program at Indian Township, Maine. The original manuscript was later edited by David A. Francis and Robert M. Leavitt for publication under the title Kolusuwakonol (LeSourd 1984). Since that time, Francis and Leavitt have continued adding to the dictionary. Their work was assisted from 1996 to 2003 by National Science Foundation grants, allowing significant expansion of the database. The database may be searched on the dictionary website,

From its inception, the dictionary has been a primary textual resource for linguists, teachers and students of the language, curriculum developers, teacher educators, translators, and writers. These users have in turn shared with the dictionary co-authors the words and glosses, sample sentences, extended texts, and linguistic insights they encounter in their work. The significance of their contributions has been discussed elsewhere (see, for example, Leavitt 1985, 1992, 2001; Francis & Leavitt 1994).

The dictionary database currently has approximately 18,000 edited entries, most of which also contain example sentences showing historical and contemporary usage. Up until the commencement of the Language Keepers project, all the words and the sentences had come directly from speakers of Passamaquoddy-Maliseet through elicitation and topic-centered interviews, or indirectly from historical and contemporary texts and recordings. While invaluable as instances of language use, the sources are essentially static in form.

The Language Keepers project has made another, dynamic aspect of linguistic knowledge accessible through the medium of video-recording. The value of the recorded sessions lies in recuperation of the natural flow of the language. In part, this derives from bringing people together for the sole purpose of speaking Passamaquoddy. This generates a flow of spontaneous talk by recreating and then capturing on film a mode of social interaction which has largely disappeared outside the home. Film also makes evident the structure, rhythm and emphasis in language provided by certain words and by subtleties of timing, facial expression, posture, and gesture. What viewers see and hear, while not always reproducible in the dictionary itself, enriches their experience with the language.

In the course of developing the dictionary, native speakers were trained in skills such as transcribing, translating, and working with consultants. These skills are crucial to a film based initiative such as Language Keepers, where Passamaquoddy conversations are filmed, transcribed, translated, and subtitled in both Passamaquoddy and English.

The documentation process
Given the endangered status of the language, it is evident that “public discourse” cannot be created using a passive documentary approach. Doing so requires a process that also addresses inhibitions to group conversation; otherwise speakers soon revert to English under the linguistic pressures that currently exist. This dilemma led the investigators to a multi-disciplinary collaboration.

The documentary filming team is composed of psychologist and filmmaker Ben Levine, and language program developer Julia Schulz. Using techniques developed over 35 years of creating cultural documentaries and developing language programs in Africa, the Caribbean, and in minority communities in the United States, their approach incorporates extensive cultural background research and a group feedback process which takes advantage of the unique playback capacities of video-recording. Most recently, a longitudinal documentary on language survival in the French communities of New England (Levine 2003) caught the attention of Passamaquoddy elders, who invited Levine and Schulz to develop a film-based language revival project in their community. The Language Keepers project evolved from that request.

The project utilizes two activities not commonly found in either dictionary building or audio-visual/documentary work.

Knowing there were inhibitions to group discourse that had strong historic precedent, we felt it important to support the documenting process with special attention to the cultural significance of place and by providing a facilitator to help participants feel comfortable and speak to their fullest ability.

Since Passamaquoddy tribal identity is closely tied to the rivers and bays of eastern Maine, we chose to begin by convening conversations in places where participants had had childhood experiences; often these places also had sacred significance. We believed that these places would elicit spontaneous conversation through recollection and the sharing of enjoyable memories. This was indeed the case. The first conversations recorded concerned social food- gathering activities and stories about the “little people,” topics used by speakers to characterize differences between the generations.

These early conversations linked to place had a positive effect in creating speaker group identity. Speakers saw that they were indeed capable of conversing in Passamaquoddy for long periods (one to two hours) with little recourse to English. They also found that in addition to talking about the past (as often happened at home) they could converse about current events.

Facilitation includes a variety of small interventions necessary to support expression among people with a history of oppression aimed at language eradication. In the French community elders could not at first speak of their feelings about language loss in complete, intelligible sentences. It was necessary to prompt speakers gently, explore links between their comments and what had preceded so as to help the full sense of their words emerge, much as a psychologist might do in interpreting linked associations to release an inhibition about talking about a painful experience (Levine & Schulz 1999, 2001). In filming Passamaquoddy, the facilitator both convened the group and helped them overcome inhibitions which might threaten the conversation.

The first facilitator was Margaret Apt, a native speaker who had been the community research coordinator for the dictionary project. As a convener, she knew speakers and had an intuitive sense of who would enjoy speaking together again and what locales they might be inspired by. As a facilitator she learned to recognize obstacles to the continuous use of the language in conversation. One common obstacle is the inability to remember a word in Passamaquoddy. Under current conditions this is the moment when speakers shift to English, as others will also shift so as to avoid awkwardness. In our process, the facilitator reminds the speaker(s) that it is normal to forget a word and switch to English, but that now we are to continue in Passamaquoddy. This always works, so well in fact that some speakers developed the custom of describing in Passamaquoddy the word they were trying to remember, rather than reverting to English. This sometimes resulted in humorous moments, as when one participant asked the name for “that thing that wiggles when fishermen use it,” i.e., a worm.

At other times, the facilitator encourages shy speakers to join in, asks questions to expand or deepen a discussion, or encourages a speaker to continue when the subject matter is difficult or painful. The facilitator also plays the important role of helping establish rapport between the filmmakers and the speakers.

Post-production feedback and editing
After the conversation has run its course, usually after one or two hours, we often show the participants a brief clip from the session. This helps them see that it was a normal conversation and allay fears that it might reflect badly on them.

Next, the facilitator and the filmmakers log the tape using the time code reference numbers, making a rough and abbreviated translation of the conversation. Since the filmmakers are not native speakers, the log is a crucial step in the editing process. Native speakers then help identify the sections of conversation deemed most valuable, as not all of the conversation can or ought to be transcribed/translated. Sections where the conversation is too fragmented to make sense are not used. Those parts of the conversation that contain clear examples of how people speak, articulate expressions of a person’s point of view, whole stories, or useful cultural information will most likely be selected for transcription/translation and editing. In addition, parts of conversations that contain “new” words or unusual grammatical constructions are earmarked for further work.

The chosen pieces are then copied and given to a transcriber for transcription into Passamaquoddy. A translator provides an English version. At each stage, time code is regularly recorded as an aid in identifying subtitle location. Finally, the filmmakers create an edited version of the conversation. The DVDs produced will have menus allowing viewers to choose English or Passamaquoddy subtitles (or no subtitles) and to open the transcription and translation texts on the screen and access the Passamaquoddy on-line dictionary, all at the same time if desired.

Film as a data source for the Dictionary
The informal storytelling and conversation filmed in the Language Keepers project highlights a number of features of Passamaquoddy-Maliseet, including word choice, noun and verb inflection, the use of particles and preverbs to shape narrative flow, and speakers’ inventiveness in creating words.

Word Choice: In telling about the time she was accosted by an apparition, a speaker eloquently parodies her own self-conscious manner of walking home along the street —

elomihtestiki ‘I went prancing off’

in contrast to her headlong run to safety:

wesuwalokittiyeqasqi ‘I ran like hell back home.’


She takes exaggerated steps to illustrate the first verb; the other she accompanies with an ironic frown. Through film the viewer sees how the first verb, literally ‘walk away tapping with feet,’ takes on a humorous, self-mocking tone and learns why the speaker chose that particular word.

Forms of Nouns and Verbs: Passamaquoddy-Maliseet verbs have hundreds of inflected forms, which first-language speakers acquired through repeated hearing in a full range of contexts. In addition, each verb has many derived forms, allowing speakers, for example, to express transitive and intransitive meanings, or to denote an action as habitual or feigned. In one sequence, a woman is telling how Whale sent porpoises to gather information about the people:

Ktopostuw-… Kciksotomoniya… Kciksotuwaniya.
‘You lis-… You listen… You listen to them.’


She starts to say a verb meaning ‘listen attentively or obediently’ but rejects this in favor of an intransitive form of a different verb that places more emphasis on hearing than attending, and finally a transitive form of this same verb. Film gives second-language learners a chance to see speakers manipulate these possibilities.

Particles and Preverbs: In learning to speak Passamaquoddy-Maliseet, it is important to know how to use preverbs (initial roots) to specify meanings, and when and where to insert the various particles that add emphasis or indicate, for example, extent of knowledge or surprise or approval. A speaker talking about his father’s last illness says,

Tan-op-al nenhitahasin, Koti wot mehcine.
‘Somehow for some reason I thought, He is going die.’


Here tan-op-al combines three particles, the first meaning ‘how’ and the others indicating potentiality and vagueness, to convey the meaning ‘somehow.’ In conjunction with this, the speaker also uses the preverb nenhi ‘in some unknown way’ with the verb final tahasi- ‘think.’ As he recalls this experience, his face bears a look of reflective bewilderment.

In another film sequence, a speaker is describing the weather as seen (and heard) through the window. She comments,

Kawatqin yaq sepay api-milawiyat Oluwis.
‘Louis says the water was rough when he went out this morning.’


Here the particle yaq ‘reportedly’ indicates that the speaker’s knowledge is indirect. Meanwhile the preverb api ‘back from having gone (to do something)’ reveals that Louis has been out on the water (milawiyat) earlier this morning (sepay) and has already returned. So listeners can assume that it was he who reported the rough conditions to the speaker. In this and the previous example, conversation illustrates clearly how these particles and preverbs can be used.

Inventiveness: Creating words is a natural — one might say, inevitable — part of speaking an Algonquian language. Words are built of meaningful roots combined to express a desired meaning. For this reason it is not feasible to list “every” Passamaquoddy-Maliseet word in a dictionary. A speaker can always change a word by substituting a different root. For example, the initial wap- ‘white’ in wapaksone ‘she has white shoes’ might become pq- ‘red’ or wisaw- ‘yellow’ or ’kan- ‘old.’ The dictionary may not list all of these words, but speakers and learners can create them ad lib. When he is telling how his young son woke him up by startling him, one speaker says,

Nit-te ntoli-sitapomoq.
‘There he was with his face right in mine, looking at me’


The verb he used was not listed in the dictionary, where its entry form would have been ’sitapomal ‘s/he looks at him/her very close up.’ Nevertheless, a dictionary user could infer its meaning by comparing it with other entries beginning with sit- ‘touching, up against, adjacent to, attaching’ and ending in apomal ‘looks at him/her.’ The latter set can be found by analogy with known verbs and based on the context in which the word is spoken. Here the English side of the dictionary is helpful. Whether the speaker had heard this verb before or invented it for the occasion, the initial mystery of its meaning shows learners the possibilities the language offers for producing new words.


Beyond documentation to awareness and action
As mentioned earlier, initiating a documentation process that requires public discourse where it does not exist is a form of intervention.

In the French communities of New England, we found that showing films in French from Quebec could elicit spontaneous French speaking among Franco-Americans who had claimed they no longer spoke French. In those communities, we were able to facilitate a discussion of how people felt about language decline. Triggered by the fresh emotions of hearing their language, which had become invisible in their lives, the discussion often led to insights about family history, choices people had made, and strong feelings for change (Levine 2003). We filmed these discussions and edited programs from them which we later played back to the Franco-American audience. This created a community dialog about language (Levine & Schulz 1999, 2001).

We referred to this dialog as “waking up” a language (in this case French; hence the title Réveil, Levine 2003). Among other things, participants could see that their language loss had an historical context, that a dominant culture had suppressed their language. Earlier generations had passed on a sense of shame about the language. Understanding this, people were able to move beyond feelings of shame and inferiority and express a strong yearning for the return of their language. Realizing that they were not alone in wanting language to return, leaders emerged and people became creatively involved in finding ways to make French a visible part of public life again. Several cities, such as Waterville, Maine, experienced language revival as a result of this process.

We have already seen indications that the process described above has correlates in the Passamaquoddy community. The filming and playback process that initially supports documentation has led to increased group discourse and the emergence of new leaders for reclamation initiatives. In one case, playback of a conversation concerning childhood experiences of being severely beaten for using Passamaquoddy in grade school led to an elder’s requesting that we film a healing ceremony she would convene to reintroduce song and language to hesitant speakers. As we enter the last year of the project, we will show back completed edits and facilitate and film the discussions that these elicit.

Preliminary findings and directions for further research
To date, approximately 35 hours of conversation have been filmed on 15 occasions with approximately 50 speakers. As these are being edited, they are shown to small groups, on Tribal TV, and in other situations that participants suggest. A special teacher’s edition has been created for use by Passamaquoddy teachers and is currently being field-tested. In addition, several individuals who are “fluent comprehenders” but do not speak the language are working with the video to explore ways to package the conversations for that “middle” generation who grew up when the switch to English was taking place.

When the project is complete in 2009, it will have developed seven DVDs containing conversations with both English and Passamaquoddy subtitles, complete with full written transcripts in both languages and online access to the Passamaquoddy dictionary. In addition the dictionary will have grown with words and sentences excerpted from the conversations.

1. The approach appears to be working as intended. Many of the elders, people who are camera shy, and a variety of fluent speakers who have participated have remarked that they enjoy the process and many have participated repeatedly. They confirm that the conversations are indeed natural and spontaneous.

2. We note an evolution in the conversations. Initially they were more like swapping stories of the old days; how we were raised, what our parents and grandparents were like, how things used to be done in a traditional way. After several conversations, people feel more secure in their use of language; they move on to topics of contemporary interest and begin to talk of more difficult issues and memories.

3. We see the Passamaquoddy listening tradition in play. When an elder had finished speaking there is often a several minute silence to permit him or her the opportunity to continue.

4. There is value in a group’s engaging in an activity as a context for conversation. When playing cards or picking strawberries, for example, the practical use of language (counting points, claiming picking spots in humorous ways) enriches the linguistic value of the sessions. Activities also help avert self-consciousness.

5. Some participants report that they feel better after discussing painful experiences in the filming and feedback sessions.

6. We note signs of the emergence of “voice” in individuals and groups as they speak with more confidence, fuller descriptions, and less hesitancy.

7. As in the French communities, we see the emergence of Passamaquoddy language advocates wanting to appropriate the documentary process to create new conversations.

8. The linguistic data and insights provided by the Language Keepers films help learners of the language become more deeply appreciative of the richness of Passamaquoddy-Maliseet. The engaging presence of the speakers on screen compels attention to their choice of words, the expression of emotion, the extent of their knowledge, and the content of their sentences and narratives. The films help learners learn — and help teachers teach — how to listen to the language and what to listen for, so that new speakers can begin to think and talk more like those whose first language is Passamaquoddy-Maliseet.


Francis, David A., & Robert M. Leavitt. 1994. More lexical exploration: Passamaquoddy synonyms. Actes du 25e Congrès des Algonquinistes, ed. by William Cowan, pp. 274-284. Ottawa: Carleton University.

Leavitt, Robert M. 1985. Passamaquoddy-Malecite preverbs. Papers of the 16th Algonquian Conference, ed. by William Cowan, pp. 73-90. Ottawa: Carleton University.

Leavitt, Robert M. 1992. Lexical exploration and educational insight. Papers of the 23rd Algonquian Conference, ed. by William Cowan, pp. 263-273. Ottawa: Carleton University.

Leavitt, Robert M. 2001. Keq nit mehsi-liwihtomon? Why do you say it that way?: Using Linguistic Knowledge in Native Language Curriculum. Linguistica Atlantica, Vol. 23, pp. 145-154.

LeSourd, Philip S. 1984. Kolusuwakonol: Passamaquoddy-Maliseet and English dictionary, eds. Robert M. Leavitt and David A. Francis. Fredericton: Micmac-Maliseet Institute, UNB.

Levine, Ben, 2003. Réveil: Waking up French — The repression and renaissance of the French of New England (video, 81 minutes). Rockland, Maine: Watching Place Productions (

Levine, Ben, and Julia Schulz. 1999. Rire Ensemble ( video, 12 minutes) Rockland, Maine: Watching Place Productions (

Levine, Ben, and Julia Schulz. 2001. Nous sommes le film… (video, 20 minutes) Rockland, Maine: Watching Place Productions (

Sockabasin, Allen. 2007 Upriver Passamaquoddy. Gardiner, Maine: Tilbury Press.