OSU Libraries | OSU Home

Translating from Kalapuya: Some Reflections

August 9, 2017

Henry Zenk, co-translator with Jedd Schrock of My Life, by Louis Kenoyer: Reminiscences of a Grand Ronde Reservation Childhood, is here to explain the complicated and painstaking process he and Jedd Schrock undertook when translating the reminiscences left by Louis Kenoyer. Kenoyer was the last known speaker of Tualatin Northern Kalapuya, the language in which he dictated his memoir, describing life and recounting his childhood on[Office1]  the Grand Ronde Reservation in Oregon in the late 19th Century. Zenk and Schrock were confronted with what many may have found to be an overwhelming obstacle: they were not left with a formal explanation of the language’s grammar. What did this mean for the translation process?

----------------------

 

The subject matter of My Life, by Louis Kenoyer is a long autobiographical narrative told in Tualatin Northern Kalapuya, the indigenous language of one of the founding tribes of the Grand Ronde Reservation community. The narrative was dictated to three linguists by that language’s last known speaker, Louis Kenoyer, who died in 1937. Jedd Schrock and I undertook this task without the benefit of a proper formal description of the language’s morphology and syntax. While the three linguists who worked with Kenoyer left us an abundance of texts and word lists, they apparently devoted little attention to ferreting out subtle variations of linguistic orm, as visible notably in the many permutations of recurring elements constituting its verb prefixes. This seems a good time to pause and reflect on just how we managed to understand Tualatin well enough to translate it into English, lacking a formal description of the language’s grammar.

The main key to our translation resides in the three linguists’ field translations, given by Kenoyer himself as the linguists read back the Tualatin text they had just transcribed from him. There are some complications, however. For one, the last quarter of the narrative (60
pages of bilingual English-Tualatin text as published) lacks a field translation, being preserved only as a Tualatin-only phonetic transcript in the field notebooks of Melville Jacobs,
the linguist to whom we owe the final form of the narrative. For another, the field translations are rather free. To some extent, this reflects the fact that Kenoyer was quite comfortable using English, in which he was not only fluent but also literate (owing no doubt to a youth spent largely in government boarding schools, one of the main subject-matters of his narrative). In common with all Grand Ronde Indians of his generation, Kenoyer had also used Chinuk Wawa from childhood, a circumstance of some significance I believe. I was fortunate to have heard Chinuk Wawa from some of the last fluent elderly speakers of Kenoyer’s home community of Grand Ronde, among them Kenoyer’s step-niece, Clara Riggs. This experience has enabled me to recognize not only Chinuk Wawa words that Kenoyer used as part of his Tualatin (and there are quite a few of those), but also, what I take to be evidence of deeper influence affecting his Tualatin word orders.

The main difficulty confronting any attempt to use Kenoyer’s field translations to decipher his morphology is that his field translations are not literal. When translating his dictations as they were read back to him, Kenoyer was clearly more concerned to produce an intelligible, colloquial English, than he was to register minute differences of morphological form. What we really need is a living, fluent speaker, with whom to explore such fine distinctions in an experimental spirit. Through a process of trial and error, we might hope to tease out the nuanced meanings lent by a speaker’s selections of particular prefixes and suffixes to express particular meanings in particular contexts. While that can be quite a long drawn-out process, it leads to the most trustworthy results. If you have any such speaker for the indigenous language you are studying, take very good care of that speaker—he or she is gold! We don’t have that for Tualatin, nor for any of the other Kalapuyan languages.

What we can do, and what Jedd and I indeed did do in preparing Kenoyer’s narrative for publication, is to use the field translations of the translated parts of the narrative as a basis for translating its untranslated parts. In the process, we each developed our own working hypotheses for identifying and interpreting contrasting word-forms in Jacobs’s transcriptions (unfortunately, we have no audio of Kenoyer from which to form our own judgments). These hypotheses were informed by available linguistic work on Kalapuyan languages, primarily analyses conducted on the neighboring Central Kalapuya language. We tested our individual
working hypotheses against the untranslated text, then took them back to the translated text to see how they stacked up against Kenoyer’s field translations. This created a mutually reinforcing feedback loop, the ultimate result of which is that we must take the main credit—orblame—for the final form of the finished translation. Our independent translations of parts of the untranslated text came out looking very similar in all cases, giving us a high degree of confidence in the final product. While we must grant that an improved control of the morphology would permit a further sharpening of the translations, we feel that our translations compare favorably on that score to Kenoyer’s own field translations.

Speaking for myself, I would describe the translation process as a kind of gestalt exercise. This was most true, not surprisingly, for the untranslated sections of the narrative. Fortunately, a full 99% of the word-stems appearing in these sections were identifiable from the translated sections. But it isn’t enough to take a segment of untranslated text, and simply string together glosses (readings) pulled from the translated sections of the narrative. While we may not grasp all of the nuances conveyed by variations in the form of, for example, the verbal prefixes, we do have non-literal English glosses from Kenoyer’s translations to go with practically every recorded variation. For more frequently used forms, we have a range of such glosses. Moreover, with respect to the verbal prefixes in particular, those in Kalapuyan languages are characteristically multi-functional. A particular prefix or prefix complex typically combines reference to the subject of the verb action with its temporal and modal characteristics (tense and aspect, actuality or potentiality of realization, etc.). The decision of which gloss from the range of recorded glosses to apply to a particular form in a particular case requires close attention to the narrative context.

In some cases, it would be virtually impossible to divine the intended meaning of a form or expression, lacking clues from the field translation. For example, among Kenoyer’s many borrowings from Chinuk Wawa is the noun “pipa.” Usually, “pipa” carries the same range of

meanings in his Tualatin that it does in Chinuk Wawa: “paper, letter, document, book”; only Kenoyer uses it in some sections of the narrative with the extended meanings “reader” (that is, lesson book) and “grade level” (referring to which graded reader a student in the reservation boarding school is up to). The untranslated text segment describing daily lessons in the reservation boarding school would be very opaque indeed, had we not a preceding translated section in which “pipa” is used explicitly with reference to students’ grade levels.

The relevance of Chinuk Wawa to Kenoyer’s narrative is apparent not only with respect to borrowings used to express key concepts, as in the above example. One of the things that struck me personally is how suggestive Kenoyer’s Tualatin word orders are of Chinuk Wawa word orders: grammatical subjects precede the verb; grammatical objects usually follow the verb, but can be fronted for focus; temporal adverbs and adverbial phrases usually come clause-first, or occasionally, appear in clause-final position; there is a single universal preposition. While subject-verb-object order is equally characteristic of English, the other features are less so. By contrast to his highly uniform and predictable word orders, Kenoyer’s verbal morphology reveals many indications of inconsistency and irregularity: prefixes are usually, but not invariably present, and come with many variant forms; productive suffixes are few, and appear only sporadically. We need more comparative work on Kalapuyan to tell how unique Kenoyer’s Kalapuyan is in these respects. My own experience was that upon immersing myself in his Tualatin narrative, I began finding it surprisingly readable, a development that I am inclined to attribute to its uniform (and, significant to my mind at least, Chinuk-Wawa congruent) word orders. By the time we got to the final stages of preparing the narrative for publication, I was able to sit down with the Tualatin text and proof it without reference to the translations. This has left me in the rather peculiar position of having achieved a real feeling for this language—only please don’t ask me to explain (in too much detail, anyway) exactly just what are all those prefixes doing!

Post new comment

The content of this field is kept private and will not be shown publicly.
Member of AAUP