Hanni Woodbury on Documenting an Endangered Language

Guest post by Hanni Woodbury, author of A Reference Grammar of the Onondaga Language

Onondaga is one of many extremely endangered Native American languages that survive now mainly because of the dedicated efforts by younger men and women who have memories of learning their ancestors' languages informally from their parents, grandparents, or elders in the community, and who, as English began to become the primary language spoken in their community, have dedicated themselves to keeping their language alive by developing extensive language programs for their communities.

Why should we document an endangered language? There are many reasons, but here are the main ones:
• because a language in many ways is a store of its speakers' knowledge;
• because a language helps to express the cultural system of its speakers;
• because speakers can use their language to express their identities;
• and, finally, because teachers and students in language programs – in the absence, increasingly, of the older speakers – must depend on the secondhand record that thorough documentation provides to keep their language alive.

The data used to document a language should, wherever possible, be a corpus of natural and spontaneous speech that is created by fluent speakers of the language. Because at this time, all speakers of Onondaga are also fluent speakers of English, it is especially important to use textual materials – stories, conversations, performances of speeches, etc. – for the purpose of documentation and to minimize the use of elicited data for fear that there might be interference from speakers' familiarity with English. The goal is to describe the language from data that represent the flow of language uninterrupted by suggestions from English.

The Iroquoian family of languages divides into two branches, a northern and a southern one. The six surviving northern languages are Onondaga, Mohawk, Oneida, Cayuga, Seneca and Tuscarora. The single language representing the southern branch is Cherokee. Today's two dialects of Onondaga exist in two locations: at Six Nations of the Grand River First Nations Reserve, near Brantford Ontario, Canada and at Onondaga Nation, which is located just south of Syracuse, NY, in the United States.

I have been involved in the task of documenting the Onondaga language ever since the early 1970s. Writing the reference grammar has been the final step in this effort. Earlier work consists of a dictionary (also published by the University of Toronto Press), and numerous published and unpublished texts with their translations.

Linguists have various methods of creating a record that will help to describe a language fully. The classic method is the one I have used: it involves a great deal of fieldwork with as many speakers of the language as possible, recording and transcribing the field sessions, asking speakers to tell stories, speeches, and recording conversations among speakers, in addition to gathering detailed information that zeros in on those aspects that are especially distinctive of the language. Beyond fieldwork, it involves scouring libraries for all sorts of published and archived materials. Eventually these records are analyzed linguistically and organized in three major ways, as a group of texts with translations, as a dictionary and as a grammar.

Numerous special features distinguish the Iroquoian languages from languages that are more familiar to speakers of English, and they impact the process of documentation. Perhaps the most noticeable feature is that the Iroquoian verb is astonishingly expandable with optional prefixes and suffixes that can be added singly or in combination to the verb root, affixes that are ordered in relation to one another. For example, the simple Onondaga verb héhsaks, given as 1 below, consists of three essential elements (morphemes) that are the minimal requirements of every well formed Onondaga verb: a pronominal prefix that references the agent (the performer of the action) and – if present – the patient (the under-goer of the action), an element that names the action itself (a verb root), and a final element that marks whether the activity is static or whether it changes through time (an inflectional element marking aspect). [Note that in the examples below morphemes are separated by hyphens]. As can be seen, this minimal verb form translates into an English sentence 'he is looking for it,' a sentence with a subject and an object:

1. héhsaks
he is looking for it

The second and third examples illustrate two of a multitude of possible ways to expanded this verb by adding prefixes and/or suffixes. In example 2 a noun root -nahd- 'comb' is inserted. It specifies the nature of the object referenced in the pronominal prefix that the agent of the verb is searching for:

2. hanahdíhsaks
he's looking for a comb

The verb with an inserted noun root shows a special feature of the Iroquoian languages, a robust process called noun incorporation.

In example 3 a suffix -h-, a so-called dislocative, marks the fact that the actor is on his way to perform the action expressed by the verb:

3. hehsákheʔ
he's on his way to look for a comb

Other prefixes and suffixes can be added to a verb root to mark directionality of the action (whether the action takes place towards the speaker or away from the speaker); the speaker's certainty as to the reliability of his assertion (it's what he heard someone say, it's what he saw with his own eyes, etc.); whether or not the statement should be understood in the negative; whether the action is to be repetitive or not; whether the statement should be taken as a command; whether the meaning of the word entails a change of state or involves two people or objects; whether the action coincides temporally with another action; whether the action is on the self (reflexive); whether one of the parties to the action is benefitted by the action; and many more. All of this and more can be signaled within the boundaries of a single word.

Another feature of the Iroquoian languages is their free word order. Unlike the affixes that expand verbs and nouns which must occur in rigid order in relation to one another, the words in an Onondaga sentence can be moved around fairly freely, depending on what it is the speaker wishes to emphasize. As a comparison, think English: the dog bites the man vs. the man bites the dog – English being a language in which word order changes the meaning of sentences. In Onondaga, who bites whom is made clear by the pronominal prefix within the verb, and the differences in the two cited sentences signal that the first option is a sentence about a dog, and the second is a sentence about a man; but in each it is the dog that bites the man. These features have a direct impact on how information is organized between dictionary and grammar.

Another distinctively Iroquoian feature is the system of pronominal prefixes that form a part of every verb – recall that they keep straight who is the agent carrying out the action of the verb and who the patient in every sentence. English has sixteen pronouns, Onondaga has fifty-eight! The details of the system are two complex to describe here, but the pronominal include three series, one to reference agents, one to reference patients, and one to reference agent and patient pairs (for verbs with two animate participants, e.g., he's thinking about her). The distribution is in part the result of verbal semantics and in part grammatical. In each of the series pronominal prefixes distinguish person (1st, 2nd, and 3rd as well as inclusion, i.e., whether or not the addressee is included in the reference), number (singular, dual, plural), and gender of animate entities (one masculine gender and two (!) feminine genders).

An English dictionary and grammar divide up fairly neatly into two functions. The dictionary lists the store of words and is heavy with meanings, and the grammar describes the rules that combine sounds into words and words into phrases and sentences. An Iroquoian dictionary as can be seen from the descriptions above, is heavy on grammar, i.e., information that deals with the construction of words – recall those multiply expandable verbs – in addition to word meanings.

The many years of studying the Onondaga language in depth have been rich with the pleasure of gaining insights and making discoveries. My hope is that this book will convey to the reader the excitement, surprises, and joys that lie ahead with the study of a language as full of intricacies and puzzles as is Onondaga.

Hanni Woodbury, Ph.D., is an independent scholar who has been researching the Onondaga language since 1971.