Journal of American Indian Education
Volume 25 Number 3
DEVELOPING ENGLISH LANGUAGE SOFTWARE FOR ATHABASKAN STUDENTS
Niki McCurry and Judith Kleinfeld
This paper presents a model for developing computer software targeted to Native American students form particular language communities. We suggest ways for school districts to develop localized software, rather than relying exclusively on commercial software not designed for Native American groups. This paper first reviews the limited research literature on computer software for Native American students. Second, we discuss the reasons school districts might want to develop their own computer software. Third, we describe the process through which the Yukon/Koyukuk School District in Alaska developed computer software for Athabaskan students. We offer both this software and the development model to other schools with native American populations. We emphasize that such software is one tool in a teacher-directed language arts program. It does not replace a teacher and it does not serve as a substitute for wholistic language activities.
DESPITE AN EXTENSIVE SEARCH through the ERIC database and other databases concerning computer-assisted instruction, we have located only three other projects which developed software specifically for Native American populations.
The first project assisted four American Indian tribes--the Tolowa, Karuk, Yurok, and Hupa--in making the transition to bilingual literacy (Schools, 1985). Humboldt State University has developed an alphabet (named Unifon) which reproduces phonetically each of the oral languages of these tribes. Using the computer's type and graphic capabilities, this project is compiling a Natural Resources Dictionary for each of these four Indian languages.
Hakes (1981) has developed a computer-based curriculum unit, "Pueblo Uses of Energy", in mathematics and science for Pueblo elementary students. The unit uses a story-telling format, designed to build upon the narrative traditions of Pueblo culture. The unit presents mathematical and science concepts in terms of the everyday experience of Pueblo children. Hakes reports large positive changes in the classroom atmosphere when Pueblo children used this computer software. The children, for example, began to talk about mathematics among themselves, to figure out ways to estimate, and to pose new problems they wanted to solve.
The most comprehensive effort to create computer software for Native American students is the Indian Affiliates' (1983) project to adapt, develop, and evaluate computer-assisted material in elementary reading and mathematics. Indian Affiliates sponsored comprehensive reviews of the research literature to identify the specific problems American Indian children have with English and mathematics (Fletcher, 1983b, 1983c). They developed a computer system (known as WICAT) which incorporates some of this information. The curriculum is currently being used on four Indian reservations (Sawyer, 1986). WICAT is aimed, however, at a broader population than Indian children alone and does not use Indian cultural materials.
The Native American computer software project described in this paper also teaches formal school English but does so within a culturally relevant context. An especially innovative and important feature of the Yukon/ Koyukuk project is that the computer software was developed to address language difficulties specific to a particular Native American language community.
The Linguistic Problem
Educators have long known that many Alaska Native students, like Native American students elsewhere, speak a dialect of English. In Alaska, this dialect is known colloquially as "village English."
Alaskan school districts spend in excess of a million dollars annually to remediate school learning problems supposedly linked to village English. Yet, until recently almost nothing was known about the nature of these dialects and what impact these dialects may have on school learning.
Village English, (hereafter LE for LOCAL ENGLISH) in Alaska may actually be a complex group of LE dialects. Each LE is a product of a unique set of circumstances; in each language region, as the original fluent speakers of the Native language learned English and passed it on to their children, they unconsciously imposed some components of the phonetic, syntactic, and lexical structure of their original language on English. Thus, there is possibly one LE for each of the twenty Native languages of Alaska. Indeed, there may be even more. The LE of a particular community develops not only from the combination of English and the original language. The LE also reflects other language influences. In the Yukon/Koyukuk region, these influences include the language patterns of French nuns and Oklahoma truckers.
Techniques for dealing with bidialectal students have been well described in linguistic literature for at least twenty years. In the sixties the linguist William Lebov (1970) outlined the technique called "contrastive analysis." He suggested that a careful study of the features of the minority dialect should be made, that these features be contrasted to academic English (hereafter AE), and that systematic instruction, analogous to that used in teaching a foreign language, be given in the difference between the spoken and the target dialect.
Labov's ideas make good sense. In order to help a student master AE, it is necessary for the teacher to be aware of the differences between AE and the student's LE. Unfortunately, despite the widespread recognition of the existence of Alaskan LEs, no systematic contrastive analysis of any Native Alaskan LE had been completed prior to the project reported here. Since the Alaska legislature approved the Alaska Bilingual Education Law in 1971, which required that special instruction in English be provided to bilingual and bidialectal Native students, Alaskan teachers and students have been floundering together in a Sargasso Sea of linguistic and pedagogical misinformation.
The Instructional Materials Problem
Even if we did have good systematic descriptions of the features of the LEs of Alaska and a good understanding of in what ways, how, and why such LEs differ from AE, we would still be faced with a dearth of instructional material to deal with those differences.
A parent might ask, "Why not just use a good English grammar book and teach them all about 'good English'?" The problem with that approach is that in fact, most so-called school "grammar books" do not teach grammar at all! They teach errors. The vast majority of "grammar texts" now on the market, including the much beloved Warriner's series, offer remediation of typical errors, not actual instruction into the structure of English.
Worse, the errors which "grammar" books remediate are not those that Alaska Native students actually make. Specifically, most grammar books address dialectal deviations found in urban Black English, Country-western English, and other dialects perceived as socially unacceptable such as Georgia Cracker and Appalachian English. Such texts are replete with lessons showing the difference between "lie" and "lay" and "sit" and "set", typical usages found in the Appalachians. It is pointless to teach Alaskan Native children to avoid misusing "sit" and "set" when they don't confuse that word pair in the first place. In fact, bringing up the sit/set error probably introduces additional confusion to the speaker of an Alaskan LE. On the other hand, Koyukon Athasbaskan students do confuse "of" and "off." Unfortunately, since that is a unique feature of Alaskan Athabaskan LE, the feature is never addressed in commercial textbooks.
The problem of a lack of reliable and systematic contrastive descriptions of the LEs spoken by Native Alaskan students may be unique to Alaska; while little has been done in Alaska, a number of useful studies of English language difficulties have been done with other Native American groups (Fletcher, 1983a). But the problem of a lack of appropriate instructional material is one shared by all educators seeking to meet the needs of particular groups of Native American students. The financial realities of the market place militate against the development of such specialized textbooks.
The Cost Problem
The plain fact is that there are over 200 different Native American languages. Each may well have generated its own LE dialect, each presenting unique problems for the learning of the Academic English the students need to succeed in college or the white collar labor market. Yet the total number of Native American students is quite small in comparison to the total number of students in the country. What publisher is going to bother researching, developing, and printing a grammar text to meet the needs of a few thousand Cherokee students, much less a few hundred Koyukon students?
One of the problems with developing such a grammar text is the high cost of preparing a hard cover printed textbook for such a small number of copies. Even if the publisher is a school district or university operating on a special grant (and such grants are available from federal and state sources) and is not seeking profit, the cost of book publication is very high. As anyone who has published a book knows, even a finished manuscript is still a long way from actual publication in terms of time, effort, and money. Layout, typesetting, and printing all add to the cost of a well designed textbook. School districts and universities, even with federal and state support cannot begin to compete in terms of quality and production with professional textbook publishers. But the publishers have no incentive to produce special textbooks for such small target audiences.
The alternative has, in the past, been the development of "homemade" locally developed instructional materials which typically have been of about the same consistency of quality as homemade wine, some quite nice and lots of it terrible. What is consistent about these locally developed, locally published materials is that they usually look just that, homemade. What kind of message about themselves, their culture, and their own importance do Native American students receive when every text dealing with the mainstream culture is hardbound, and in three colors on thick shiny paper while every text dealing with Native American language and culture is hand-stapled, in purple mimeo, and on xerox paper?
While printed books are not feasible to publish for small target populations, computer courseware is cost-effective. Both the courseware developer and the writer must research the content area, develop an instructional plan, and design educational activities. The book must be written and the computer disk programmed. But here the two processes diverge. At this point, the computer disk is done, it only needs to be copied to be "published" and disseminated. The writer must find a publisher, send the book off for typesetting, layout, proofing, and printing. All this costs time and money.
Not only are computer disks easier to publish than a professional printed book, but there is no discernible visible difference between a commercial computer disk and a locally developed one, provided both are equally well designed with regard to content, graphics, sound, and so forth. It makes a good deal of sense for a school district serving a particular population of Native American students to develop its own courseware. Yukon/Koyukuk has found the cost of developing localized software to be about $5000 for one hour's worth of instruction. The more students who use the disk, the more cost effective the disk is. If 100 copies of the disk are used, then the cost per disk is $50. This is no more expensive than the cost of a piece of commercially developed software and is instructionally more appropriate to the needs of a special group.
The process is logical, inexpensive, and based on sound educational theory. Why haven't more educators done it? As in any new field, a model is needed.
A Model For Culturally Specific Courseware Development
This paper presents a practical model for how a school district can develop its own software to meet the identified needs of a unique student population. We developed this model through PROJECT APEL (APPLES for PROFICIENCY in the ENGLISH LANGUAGE), a Title VII funded project which focused on using Apple microcomputers for improving English language skills of 300 Koyukon Athabaskan students in ten villages in Interior Alaska. Since 1982, APEL has published and disseminated 13 computer kits offering more than 100 hours of instruction in English grammar and the writing process.
The APEL courseware development process has essentially six steps: 1) a contrastive analysis of LE; 2) establishing instructional priorities; 3) identifying appropriate commercial software; 4) selecting objectives for software development; 5) developing the software; and 6) implementating the software through staff training and dissemination.
Step 1: Contrastive Analysis of LE Dialect
Our primary source of information about the LE dialect is a student writing sample. We also use teacher observation, national standardized tests, and linguistic research to identify the distinctive language difficulties of our students.
Each year, the writing sample assignment (called the "prompt") is designed by a team consisting of the APEL instructional coordinator, a linguist (Athabaskan specialist) from the University of Alaska Native Language Center, and the researcher who will conduct the detailed analysis of the writing samples. The group decides which category of information is of special interest for that year, and then designs an assignment which is likely to elicit that type of miscue from the students. For example, the FY86 sample will focus on tense usage, so a narrative or story assignment is planned. The topic should be one on which students grades 4-12 can write with some assurance; so, it should relate to their own communities or experience. Students in grades K3 are not sampled since the writing of young children is likely to show immature developmental idiosyncrasies not necessarily related to dialect difference.
The team writes a prompt based on the selected topic. The prompt includes a student assignment page and teacher instructions. Teachers administer the sample assignment on the same week. Students write for up to 40 minutes and are given the opportunity to edit using a dictionary or grammar handbook if they ordinarily do so as part of their usual writing activity in class. Teachers collect the samples and send them unmarked to the researcher.
A key to the success of the analysis is the background of the individual researcher. The researcher must know the types of errors likely to be of interest to teachers and must have a good linguistic background in both English and the original Native language. The researcher analyzes each essay, counting errors and categorizing the errors by student's grade level, gender, and village.
The resulting research report includes the data charts of errors, an interpretation of the data, and recommendations for instructional application. The interpretive report includes not only a list of the most frequent errors or patterns of error, but divides the error categories into 1) errors made by most students nation-wide; 2) errors typical of any English as a Second Language (ESL) speaker of English, and 3) errors specific to that particular Native language group. In the case of the latter, the report explains any possible source or motivation for that error from the original Native language or distinctive historical pattern of culture contact. APEL has generated an annual report of the writing sample analysis each year since 1982. In 1985, a three-year summary report was published (McGary, 1985).
Step 2: Establishing Instructional Priorities
The contrastive analysis provides the raw data against which teacher information and standardized test scores can be analyzed. We use the individual item analysis report provided by the Iowa Test of Basic Skills scoring service. We cross-check the interpretive data regarding the relationship between the original Native language and miscue patterns with an independent specialist in the original language of the region, Koyukon Athabaskan. A published grammar of Koyukon Athabaskan provides an invaluable resource.
At this point, we move from an identification of error and dialect miscue patterns to establishing instructional priorities. Crucial to this prioritization process is a clear knowledge of the end goals or final student outcomes sought. For APEL, we established that our final desired outcome was that the student should have an ability to write AE so that after high school graduation, the student could enroll in a freshman English class at a typical open admissions university or common college and not immediately flunk out. A team of teachers and linguists generates the prioritized instructional objectives list.
From the prioritized list a number of useful documents are produced. First, a summary report lists these prioritized objectives, their relationship to the original Native language, and the instructional implication of that relationship. Second, the objectives are added to the district curriculum scope and sequence, thus becoming part of the expected learning in that subject. Third, the objectives list can be used as a materials needs assessment checklist for selecting or developing computer materials for the APEL project.
Step 3: Identifying Appropriate Available Software
Since it is much less expensive to purchase commercial software than to create our own, we first identify existing material that serves our objectives. Using the prioritized list, the instructional coordinator and the district medial specialist review commercial computer software which purports to teach the targeted objectives. We buy one copy of the program, and a teacher on site pilots the material with our students. Software which passes this selection process is then purchased for all APEL sites and disseminated.
Step 4: Selecting Objectives for Software Development
This sifting and winnowing process results in a list of instructional objectives in two categories: 1) objectives for which no software exists at all, such as specific Athabaskan-English homophone miscues (of/off) and 2) objectives which existing software deals with inappropriately, such as the verb tense system which differs so significantly in LE and AE that even programs such as Milliken's Verbs are inadequate.
Step 5: Developing Software
At this point we know what we want to teach and can specify design requirements. Each computer package addresses one instructional objective through a tutorial which provides both instruction and practice in applying the principle. These learning activities occur in the context of an interactive adventure or fantasy taking place in a rural Alaskan Native village, with activities familiar to the students, such as dog mushing, gold mining, or moose hunting.
A serious problem in developing special software is finding someone to do it. The supply of experienced educational software programmers familiar with the lifestyle of our Native Alaskan students is limited. Fortunately, software authoring languages are designed to fill the gap between the educator and computer programmer. Designed specifically for educators who want to write software, they are already programmed with short cut instructions for accepting right or wrong answers, animation, and sound effects. We selected APPLE SUPERPILOT, a program which, unlike the real computer languages such as BASIC or PASCAL, can be learned in about a week. This option widened our choice of developers from a few programmers to a number of experienced educators.
We advertised for computer kit developers among our staff and among experienced educators in Alaska. We paid a fixed sum per kit, based on the assumption that a kit would take a developer about one month of full time work to complete. To get the teams started, we gave a week-long seminar which covered instructional software design, SUPERPILOT, and linguistic information about LE. We required each developer to come to the seminar with a disk design including the instructional objective for his/her kit (such as mass/count nouns), the topic or theme of the kit (such as sled dog racing), and an instructional technique (cloze, multiple choice, sentence combining). The co-operative seminar atmosphere, we have found, produces better graphics and story lines and results in software oriented toward higher order thinking skills more than when teams are working in isolation. Because of the disparate topics covered, the seminar was team-taught by a linguist, a media design specialist, and a bilingual educator.
Step 6: Software Dissemination
Instructional materials must always be accompanied by training. Otherwise the materials are never used or they are used incorrectly. Each APEL kit includes one or more computer disks and a hard cover teacher guide which specifies the instructional objective, the prerequisite skills (such as the reading level required), the hardware required (such as a printer or color screen), appropriate instruction to be given prior to use of the software, and appropriate follow up activities. APEL software, as is true for any educational software, is not intended to replace teacher instruction, but to supplement it; APEL grammar instruction is meant to be a part of a total composition program, and not to comprise the entire program. Each year, a one-day workshop is given at each APEL site to introduce staff to the appropriate use of the newly developed APEL software.
Results of Courseware Development
We have found the materials development process outlined above to be time and cost effective. In just over three years and for less than the cost of the average commercial courseware package, we have developed thirteen kits offering hundreds of hours of instruction. Our APEL kits have not only been disseminated to every classroom in our district, but to more than half of the schools in Alaska, including those in the two largest cities, Fairbanks and Anchorage. The kits are being used; we get calls and comments about them from all over Alaska. The materials have been found to be of interest to all Alaskan Native students, not only Athabaskan students, because they are professional looking, attractive, and treat in knowledgeable detail the lifestyle of the students. Indeed, Iditarod Musher, featuring the celebrated annual long distance dog sled race, is a popular favorite among all Alaskan children, regardless of ethnic background.
The kits are fun and attractive, but do they teach anything? We do know that, since the APEL project began, student reading and language usage scores on standardized tests have crept upwards on a steady sustained basis, from average national rankings below the 10th percentile to just below the 40th percentile. We cannot attribute such gains solely to the kits. The kits are only part of the instructional improvement process APEL fueled.
We do know that, in design, the kits are equal to commercial material, that they interest the children, that they teach content which we know must be taught, and that they validate and give our Native American students pride in their own lifestyle.
Our APEL kits are available to any school district at a minimal cost. Districts serving Athabaskan students such as Navajo or Apache might find our kits of direct use. Districts serving Native American students from other language groups might wish to adapt our development model. We have found the process to be an exciting, cost-effective and energizing project for staff and students alike.
Appendix I: Developing Software
If you decide to develop your own computer software, this is what you will need to get started:
Appendix II: APEL Kits Available
APEL KITS are available for purchase to any non-profit educational institution. With the kit, purchaser will receive a license to copy within district or school; all programs are copyable using CAPYA program on the APPLE System Master diskette. All kits include a teacher guide booklet.
*Includes extensive manual.
Fletcher, J. D. What Problems Do American Indians Have with Mathematics. Report prepared for Indian Affiliates, Orem, Utah, 1983a (ED 247052).
Fletcher, J. D. What Problems Do American Indians Have with English. Report prepared for Indian Affiliates, Orem, Utah, 1983b (ED 247051).
Fletcher, J. D. A Bibliography of Studies on Elementary and Secondary School Reading, English, and Mathematics for American Indian Students. Prepared for Indian Affiliates, Orem, Utah, 1983c (ED 2470050).
Hakes, J. A. Computer Storytelling Mathematics for Pueblo Indian Upper Elementary Level Students. Final Report, Washington, D. C.: National Institute of Education, 1981 (ED 215888).
Indian Affiliates Inc. Computer-Aided Instruction in Education Basics for Indian Students. Final Report, Orem, Utah: October-June, 1983 (ED247054).
Labov, W. The Study of Nonstandard English. Urbana, Illinois: National Council of Teachers of English, 1970.
McGary, J. Analyses of Student Writing Samples from the Yukon/Koyukuk School District 1983-1985. Nenana, Alaska: Yukon/Koyukuk School District, 1985.
Sawyer, T. Indian Affiliates, Orem, Utah, Personal Communication, 1986.
Schools, The leap to literacy, Electronic Learning, Nov.-Dec. 1985, 9-10.
[ home | volumes | editor | submit | subscribe | search ]