|
Article Excerpt
ABSTRACT. This article represents an additional contribution to the steadily growing body of work that addresses the problems faced in creating dictionaries for Native American languages. In his 1975 Cherokee-English dictionary, Feeling cites verbs in the third-person singular conjugated forms. Cherokee is a polysynthetic language in which all verbs are bound roots; accordingly, bare verbal stems do not surface for speakers. Today's potential audience differs from the 1975 audience in a number of ways, however, which render the choice to use conjugated forms less appealing. This paper asserts that listing verbs by their unprefixed stems is a linguistically sound alternative which will make the dictionary more helpful for those individuals most likely to use it. Three arguments in favor of using the unprefixed stems as citation forms are presented. First, complex phonological rules render certain forms unpredictable based solely on the conjugated forms given in the dictionary. Second, dialectal variation causes surface forms to vary; using a common underlying stem will make the dictionary more accessible to different speakers. Third, using stems will place derivationally-related words together and increase the dictionary's use as a pedagogical tool while involving a relatively minimal amount of parsing. The paper concludes with suggestions for making a practical electronic dictionary of a polysynthetic language that is useful for a new generation of L2 users as well as linguists. *
1. INTRODUCTION. The purpose of this paper is to offer a possible solution for problems involved in the creation of a Cherokee-English dictionary. The problems a lexicographer faces when creating such a bilingual dictionary are typical for many languages of the Americas. Mithun points out that many Native American languages are polysynthetic; that is, they are characterized by words with a high number of morphemes (1999:38). Verbs in these languages frequently have obligatory inflection for person, tense, aspect, and other grammatical features. The lexicographer of a polysynthetic language is faced with the dilemma of either using an inflected form as a citation form or using an unnatural root form. This paper examines the dictionary of one such polysynthetic language, Oklahoma Cherokee, and suggests that an unprefixed stem form is more appropriate than a naturally occurring form. I will present three arguments for using the stem. First, the prefixed form often obscures the underlying form of the stem and thereby makes other prefixed forms unpredictable. Second, prefixed form entries create difficulties for speakers of different varieties of Cherokee. Third, using the stem form can serve as a learning tool by placing derivationally-related words together. While this format presents some difficulties of its own, this paper will argue that such a format is better suited to the individuals most likely to use the dictionary, namely, L2 users who have been exposed to the language in a classroom environment. Furthermore, most of the potential drawbacks of such a design can be diminished by using an electronic format. While there are valid concerns regarding the use of stems as citation forms, the increasing ability to translate paper dictionaries into searchable, electronic databases may provide excellent solutions to very real problems.
1.1 PREVIOUS WORK ON CHEROKEE. The Cherokees are one of the largest and most-recognized indigenous groups in the United States. Given the size of this group, it is surprising that there are relatively few pedagogical or linguistic materials available for the language. Several sketches and articles were published in the first half of the 20th century (Bender and Harris 1946; Reyburn 1953a, 1953 b, 1954), and two dissertation grammars of North Carolina Cherokee appeared in the 1970's (King 1975, Cook 1979). In the following decades the most important contributions were a dissertation on grammatical relations and verb agreement (Scancarelli 1987), two chapter-length grammatical sketches (Scancarelli 2005, Walker 1975), a collection of UCLA linguistic articles devoted to Oklahoma Cherokee (Munro 1996a), and a Cherokee-English Dictionary with an attached grammatical sketch (Pulte and Feeling 1975). These last two works, as well as my own field work, form the basis for much of the discussion here. In the 1980's Geoffrey Lindsey (1985, 1987) produced two important papers on the role of tone in Cherokee phonology; the second of these is unpublished, but has important insights that several of the UCLA working paper authors use in their analyses. Besides these linguistic resources, there are a number of learning materials of varying size and quality; the largest and most useful is Holmes and Smith's Beginning Cherokee (1977).
1.2 THE SOUNDS OF OKLAHOMA CHEROKEE. Cherokee has a relatively small inventory of sounds; most of the consonants (with the exception of the lateral affricates /dl/ and /tl/) are familiar to a speaker of English or other European languages. Cherokee has six major vowels /a, e, i, o, u/, as well as a mid central vowel with slight nasalization represented as . Vowels can be long or short. Cherokee is a pitch accent language with two level tones, low and high, and four contour tones: rising, falling, lowfall and highfall. (1) The traditional Cherokee writing system does not reflect these suprasegmental features; this paper will represent vowel length using a single vowel to represent a short vowel and a double vowel to represent a long vowel. Tone is represented using the system delineated in (1). All material cited herein is adapted to this orthography.
(1) Cherokee vowels: length and tone e short low ee rising ee long low ee falling e short high ee lowfall ee long high ee highfall
Tone will be significant in this paper only with regards to the lowfall. In Oklahoma Cherokee, the lowfall often results from the insertion of a glottal stop; this glottal stop is realized in this dialect as a lengthened vowel with lowfall; e.g. /e?/ becomes /ee/ in Oklahoma Cherokee.
It is crucial for the discussion that follows to point out that the sounds represented in the dictionary's orthography (as well as the orthography in this paper and the UCLA papers) as , , and are actually the unaspirated voiceless stops /t/, /k/ and /c/. Their aspirated counterparts in the dictionary's orthography are , , and (sometime ), phonetically /th/, /[k.sub.h]/ and /[c.sub.h]/(sometime/[ts.sup.h]/). In many linguistic works on Cherokee (King 1975, Cook 1979, Scancarelli 1987) the unaspirated stops are represented as and and their aspirated counterparts as and .
1.3 THE CHEROKEE ENGLISH DICTIONARY. The phonological analysis that will be used in this article is based on the important work done by Pamela Munro and her students in the UCLA papers. These insights are applied to the 1975 Cherokee-English Dictionary which was written by Durbin Feeling, a native speaker, and edited by William Pulte. Pulte and Feeling collaborated on a grammar sketch; these two works appear in a single volume. (2) Feeling states in the dictionary's introduction that it is intended to serve three types of users (Feeling 1975:vii). First and foremost, it is meant to serve Cherokees who already speak their language by improving their knowledge of the syllabary as well as helping them understand the grammatical structure of the language. While native speakers are the primary audience, the author makes it clear that his work is also intended for students learning Cherokee, as well as linguists who wish to gain information about the structure of the language. For these last two groups, one of the most appealing features of the dictionary is the listing of example sentences (at least one and sometimes two) for every entry. These sentences appear first in the syllabary, followed by a Romanized script and a natural English translation. The Romanized script used in these example sentences is distinct from that used for the entries and subentries in that it omits indications of vowel length and tone. It is crucial to keep these distinct audiences in mind, as they directly affect choices regarding organization and format. This work has proven immensely useful to both linguists and non-linguists for several reasons. The most striking feature is the concurrent use of the Cherokee syllabary and the Romanized alphabet. The dictionary is alphabetized according to the Romanized characters, with the syllabary forms appearing next to the Romanized forms.
The Romanized script is useful for linguists in part because of the addition of the superscript features indicating tone; it should be pointed out, however, that the presence of superscript numbers inside of words is probably the most daunting feature of the dictionary for non-linguists. In addition to these numbers, a dot under vowels indicates vowel shortness. A sample entry is in (2).
(2) [TEXT NOT REPRODUCIBLE IN ASCII] [gu.sup.2][hi.sup.23][sdi.sup.3]ha v.t. he's...
|
