DOCUMENT RESUME

ED 276 294 FL 016 265 AUTHOR Cowan, J. Ronayne; Cziko, Gary A. TITLE Evaluation of the Southern Sudan Local Languages Literacy Project: Final Report. INSTITUTION Center for Applied Linguistics, Washington, D.C. SPONS AGENCY Agency for International Development (IDCA), Washington, DC. PUB DATE Nov 84 NOTE 235p. PUB TYPE Reports - Evaluative/Feasibility (142)

EDRS PRICE MF01/PC10 Plus Postage. DESCRIPTORS *African Languages; Arabic; *Developing Nations; Educational Attitudes; Elementary Education; English (Second Language); Foreign Countries; *Language of Instruction; Language Research; Language Usage; *Literacy Education; *Native Language Instruction; Program Evaluation; *Technical Assistance; Uncommonly Taught Languages IDENTIFIERS *Sudan (South) ABSTRACT This report describes an evaluation of the Local Languages Literacy Project in Southern Sudan, a project to develop literacy education and instructional materials in the various vernaculars of that region and to create a bridge to the more common English and Arabic usage. The first chapter gives background information about the project and the involvement of the United States Agency for International Development and the Center for Applied Linguistics. Chapter 2 examines the social context of the project and describes the research concerning local language and educational attitudes and language use. Chapter 3 outlines the development of the literacy materials at various levels, including those developed for English and Arabic instruction. Chapter 4 describes three evaluations of the project's local impact and discusses the results. Chapters 5 and 6 summarize the major lessons learned from the project. A list of references and instructional materials is appended. (MSE)

*********************************************************************** Reproductions supplied by EDRS are the best that can be made from the original document. *********************************************************************** Evaluation of theSouthern Sudan

Local Languages LiteracyProject:

Final Report

J Ronayne Cowan andGary A. Cziko

University of Illinoisat Urbana4hampaign

and

Center for Applied Linguistics, WashingtonD.C.

iloveaber 19811

U.S. DEPMTMENT OF EDUCATOR Office of Educational Research and Improvemmt "PERMISSION TO REPRODUCE THIS EDUCATIONAL RESOURCES INFORMATION MATERIAL HAS BEEN GRANTED BY CENTER (ERIC) ATMs document has been reproduced as received from the person or organization wigimtinctit. 0 Minor changes have been made to improve reproduction Quality. ,yadeA, Points of view or opinions stated in this docu- mentdo notnecessarily represent official TO THE EDUCATIONAL RESOURCES OERI position or policy. )INFORMATION CENTER (ERIC)."

BEST COPYAVAILABLE PAGEL

CONTENTS

Preface iii Acknowledgements v 1.0Background to the USAID Project and the CAL. Evaluation 1 2.0 Social Context of theProject 2.1 Introduction 6 2.2 6 Description of Research . 2.2.1 Research Design 7 2.2.2 Data Collection 7 2.3 Results . 7 2.3.1 Background Data 9 9 2.3.1.1Misir . 2.3.1.2 Mayom 9 2.3.1.3 Tambio 10 . .. .. -11 2.3.1.4 Naga . 2.3.2 Questionnaire Data 11 12 2.3.2.1Language Use and Attitudes 12 2.3.2.2 Urban-Rural andAgrarian-Pastoral Differences 2.3.2.3 Reading Interests 13 14 2.3.2.4 Availability andControl of Education 2.3.2.5 Teaching Methods 14 15 2.3.2.6Importance of Education 16 2.3.2.7 Consequences of Education 2.3.2.8 School Attendance 17 2.4 Discussion 18 19 3.0 The Literacy Miterials 3.1Method 24 3.2 Design of the Materials 24 3.3 Problems with the Cycle 25 1 Materials 26 3.3.1 Problems with Orthography Construction 27 3.3.2 Problems Related toVocabulary Control 3.3.3 Proofreading 30 3.4 The English Bridge 33 Materials 34 3.5 Recoamendations for Improvenents 3.6 The Reading Workshop 35 36 3.6.1CUrrent Reading Theory andResearch 3.6.2 Innovations in the Literacy Materials 37 41 3.6.2.1Cycle 2 Innovations 3.6.2.2 Cycle 3 Innovations 41 3.7 Cycle 2 Orthographies 44 3.8 The Arabic Materials 45 45 4.0 Impact of the Project 48 4.1 First Impact Evaluation 49 4.1.1Description of Evaluation 49 PAGE ii

4.1.1.1Evaluation Design 49 4.1.1.2 Instruments 4.1.1.3 Data Collection 49 4.1.2 Results 50 4.1.2.1 Background Data 51 4.1.2.2Process Data 51 54 4.1.2.3 Outcome Date 4.2Second Impact Evaluation 55 57 4.2.1 Description of Evaluation 57 4.2.1.1 Evaluation Design 57 4.2.1.2 Instruments 4.2.1.3 Data C011ection 58 , 60 4.2.2Results 61 4.2.2.1 Background Data 62 4.2.2.2 Process Data 63 4.2.2.3 Outcome Data 67 4.3 Third Impact Evaluation 79 4.3.1 Description of Evaluation 79 4.3.1.1Evaluation Design 79 4.3.1.2 Instruments 79 4.3.1.3 Data Collection 4.3.2 Results -81 82 4.3.2.1Background Data 82 4.3.2.2 Process Data 85 4.3.2.3 Outcome Data 4.4Discussion 87 4.4.1 96 Primary Evaluation and Literacy inthe Rural Southern Sudan 96 4.4.2 Overall LiteracyAchievement 97 4.4.3Evidence of Impact of the Project 4.4.4Implications 98 99 5.0 General Discussion 102 5.0Postscript 111 7.0References 113 8.0 Appendixes 119 PAGE iii

PREFACE

This report describes, insone detail, a reasonably rigorous investigation of diverse aspects of the LocalLanguages LiteracyProject conducted in Southern Sudan. In numerous areas of theworld, educators are faced with necessityto make important the decisionsconcerning the choice andoptimal sequencingoflanguages for initial literacytrainingand forcontent instruction. Southern Sudan is an area of extremelinguistic heterogeneity in whichliterally dozens of mutuallyunintelligiblelanguagesarespoken. Arabic isthe official language of thecountry although Englishoccupies a special position as a lingua franca inthe southern region, thelanguage of the regional governaent, and the language of communicationbetween Sudan and non-Arabic-speaking states. Therefore,high premium is placedon developing proficiency in both Arabic and English to facilitate educational,occupational and social mobility.

This task although complex ismorestraightforward in the northern part of the country where Arabic is widely spoken as a mothertongue. In the South, the numerous mutually unintelligible languages are so unrelated to Arabicthat one canlegitiaatelyraise the question of whetherachild initially introduced to literacy training and tocontentinstruction in the home language might, over.the longrun, benefit lore from formal instruction thana counterpartwho is *submerged" in Arabicfromthe beginning of formal schooling. For our purposes, then, a aajorfocus of theproject concerned whether one could optimallychoose andsequence languages for purposes of initial literacy training and initialcontent instruction to maximizethe educational benefit derived by participating children and thus helpthem to enhance their potential.

Analogous situations exist and thedilemma isrepeated innumerous countries throughout the world-- in the Philippines, in Nigeria, in Cameroon, etc. -- and hence the results of this investigation will interest not only Sudanese educators but the myriadothers throughoutthe world whoare also concerned with such questions.

Initially, a very optimistic research agendawasenvisioned. There would be stages for needs assessments, planning,materials preparation, teacher training, trial implementation, andrevision followed by the careful selection of groups of project and comparisonchildren who wouldbe observedover a period of years during their primary school training andthe introduction of revisedmaterials.The plan called for an examination ofchildren from. several ethnolinguistic groups from the time they began their formalschooling and were or were not introduced to initial literacy using newly developed vernacular materials to the time whenLley undertook a transition toArabic as the medium of instruction and also beganthe study of English asa second language. Ideally, the childrenwouldhavebeenfollowedthrough the completion of the primary cycle of their education-- i.., Primary 6. By then, it would have been possible to look cumulatively at theprogress made by the children in the comparison groups and in the project groups. One would be ableto exasinethe children's languageabilities, their contentsubject

5 PAGE iv

mastery, their school retention rates, theirperformance on the primaryschool leaving examination and their relative ratesof admission to junior secondary school. Other affective informationcould havebeen examined aswell. Unfortunately, despite the fact that theprogram began with a flurry enthusiasm and with a good of dealofsolidaccomplishment inmaterials development andteacher training, workdid notproceed as rapidly as might have been hoped because of various social, economic andpolitical factors. These have affected not only the educationalsystem but also the social fabric of SOuthern Sudan during the past fewyears.

The present report, then, is a modest one, but nevertheless in my view an important one. It is important in thatit presents a aodel for consideration by edueationalresearchers in othersimilar settings-- and there are many throughout the world. It is important also because itraises explicitly the question of the utility and benefit to be derived from local languageliteracy projects in settings in which theta is virtually no support whatsoeveroutside the formal school context forsuch literacy once acquired. Theproject is also important because of thedemonstration that such innovation necessitates teamork involving people with training inanthropology, linguistics, and psychology together with those committedto educational reform. The monograph describes the efforts of sucha group. Southern Sudan is a difficult place.in which to work, and an even more difficult place in which to conduct"rigorous" research; butthe report demonstrates that it is possible to ask meaningful questions in e rigorous fashion insuch settings.

G. Richard Tucker, Center for AppliedLinguistics PAGE v

ACKNOWLEDGEMENTS

It is virtually impossible to acknowledgeall those individuals whohelped tomake this report possible. These individualsrange fromthose at the United States Agency for International Development whose funding madeboth the literacy project and this researcha reality to the teachers, children in parents, and Southern Sudan who were involved in this research. Among these individuals are: UshariMahmud and the late James Dahab Gabjanda ofthe Institute of African and AsianStudies of the University of Khartoum; O'Brien of the Lutheran World Michael Federation; SejarioLatansio, Massimino Allem, Sullen Nginzo, DavidTenni, Simon KuoonPuoc, EdwardManderson, andJob Dharuai ofthe Institute of Regional Languagesof the RegionalMinistry of Education in Juba, Sudan; John Hollman, Julie Van Dyken, Rick Brown,Joycelyn Clevenger, Alice Van Bergen, RichardBergman, Wanda Pace, Robert Hoppe, David Bendor-Samuel, and John Bendor-Samuel of the Summer Institute ofLinguistics; and G. Richard Tucker of the Centerfor Applied Linguistics. These latter three individuals deserve special recognition for their carefulreading of earlier drafts of thisdocument andfor theiruseful comments and written contributions to it. Others also deserving acknowledgementareEluzai Moga Yokwefor hisassistance in thedevelopment ofthe Bari-language testing materials andfor the analysis of the Barioral reading errors, and Marian Farquhar whose hospitality and in-depth knowledge of Nuer languageand culture greatly facilitated the workdone in Emir.

We also wish to thank the school directors and teachers who participatedin this research, all of whom gave usethe fullest cooperation in allowingus to observe classes and obtain informationon the schools, teachers, and pupils. Finally, we wish to offer special thanksto the primaryschool children theaselveswho providedus with fascinating and usefuldataand who, we sincerely hope, will ultimatelybenefit from our work.

7 PAGE 1

1.0 BACKGROUND TO THE USAID PROJECT AND THE CAL EVALUATIJN

In Novesber 1975 the High Executive Council Jf the Southern Region of Sudan passed Resolution 0273 which laid down basic Gducational policy as follows:

.The Use of Local Languages and of Arabic and English forEducation in the Southern Region.

After considering submissionof the Regional Minister of Education No. MESR/SR/1. A. 1/5 dated23-10-75 on the use of theabove languages for education in the Southern Region in the light of its Resolution No. 247 dated 29-9-75;

The Hon. H.E.C. in its meeting No. 103 of 8-11-75 resolved that:

a. in the case of rural schools:

i) The Vernacular be used as medium of instruction in the first and second years with Arabic and English introduced orally;

ii) The Vernacular be used as mediumof instruction in the third and fourth years while Arabic and English are intensified.

iii) Arabicbe the mediumof instruction in fifth andsixth years while Englishcontinue to be intensified.

In taking thisdecision the High Executive Councilwas deliberately going back to the situation whichexistedbefore the civil war whenprimary education began in the local languages. It was the understanding of the High Executive Council that theAddis AbabaAgreement of 1972, whichended the civilwar, gave the South the right to use their local languages. That agreement had led to the establishmentof a semi-autonomous government in the Southern Region with its ownregional sinistries, cabinet (HighExecutive Council) and president of that council.

One of the early policy decisions of the Regional Ministry of Education was to take step3 tolook into the language situation and the Ministry asked the Summer Institute of Linguistics (SIL) to assist it in this. As a consequence, SIL carriedout alanguage survey in the period October 1974 toMay 1975. This survey identifiedninelanguages (Bari, Dinka, Kresh, Lotuho, Moru, Ndogo, Nuer, Shilluk, and Zande)spoken by the largest ethnicgroups which could be used as media of instruction in primary schools. As pan be seen from Table 1.1 which shows the estimated population of the languages identified in the survey, these nine languageswere thought to comprise some 2.7 million speakers.Anadditional 14 languages (Acholi, Anuak, Baka, Banda, Didinga, Ferroge, JurBeli, Jur Luo, Kakwa, Kaliko, Mundu, Murle, and Taposa) were also identified as needing literacy materials so that children could at least learn toread and write in them even though anotherlanguage might be used for instruction in school. The nine RoleA languages, as theywere PAGE 2

designated, and the 14 Role B languages were estimatedto embrace over three millionspeakers. The survey also found the rateofbilingualism in the Southern Region to be such that an additional 225,000 speakers in othergroups could also be servedby materials produced in the RoleA andB languages. This would encompass about 3.3 million individualsor 95% of the estimated population of the region. Four more languages (AvUkaya,Belanda Bor, Belanda Bviri, and Mabaan) were added to the RoleB group in 1980.

Following this survey, the High Executive Council made its decision touse the local languages in the rural schools, i.e., the schools outside thethree towns of Juba, Mau, andMalakal, by passing Resolution 0273. To implement this decision, the Southern RegionalMinistryof Education took three steps. On 15January 1976, the Ministrysigned acooperative agreementwith SIL. this agreement states:

Whereas the Ministry and the Institutehave expressed theirdesire to cooperate ina project designed tocommence formaleducation in the southern region through initial literacy in themother tongue followed by graded transfer to the major-languages. 1. The Institute shall (1) (a) in case oflanguages where basic linguistic research has already beenundertaken train personnel selected by the Ministry to write thenecessary teaching materials and test them; As a second step, later that year the Ministry set up a department ofLocal Languages Development. This functioned, at first,as a department within the regionalMinistry. A yearlater the departmentbecame the Instituteof Regional Languages (IRL).

The Ministry's desire to use the local languages as soon as possibleled to anotherstepthatyear. Oneresultof the civilwar had been the disappearance of copies of primers in the local languages, which had beenused before the disturbances. Wherever copies ofold primers in local languages could be found, the Ministry had themreprinted. This proved to be a popular decision. Southerners wanted their languagesrecognized and in use as quickly as possible and this was seen as the first stepin bringing this about.

During 1976 and thefirst part of 1977, the Ministry ofEducation and SIL discussed plans for a joint project. Suitable teachers would be trained to write new primers in the nine major.languages using the existingorthographies ofWhich the southerners were proud. Any ideaof revising theexisting orthographies was conceived as likely to arouse controversy and to delay the process of writing new primers. Sincethe aajority of southern officialshad learnedto read and write using theexisting orthographies, this wasnot surprising.

The respectiveroles of the Ministry andof SIL were spelled out. The Ministry would provide the teachers and the facilities for the training. SIL would provide staff to give thetraining and to undertake general supervision of the Whole process of writing the new primers, testing them, revising them and teaching teachers in theschool system to use them. Itwas recognized that the high degree of person-to-person contact between theSIL staff and the Ministry of Education personnelwhich would be necesseary tomake the training and the subsequent writing of thematerials effective, made itadvisable to PAGE 2a

Table 1.1

Estimated Population of Regional LanguageGroups in Southern Sudan in 1978

Language Population

Bari 226,000 Dinka 1,275,000 Lotuho 132,000 Ndogo 25,000 Moru 68,000 Nuer 525,000 Shilluk 124,000 Zande 236,000 Kresh 16,000 Murle 51,000 Jur Luo 54,000 Jur Be li 22,600 Mundu 10,000 Acholi 27,000 Anuak 14,700 Baka 23,000 Banda 10,200 Didinga 58,000 Ferroge 5,600 Kakwa 40,000 Kali ko 7,000 Madi 18,000 Toposa 138,000

3,106,100

Note: It is estimated that an additional 225,000bilingual speakers of other language groups could be served by materialsin the above languages bringing the total to 3,331,100.(Source: Operational Grant Proposalto USAID by the Southern Ministry of Education and theSummer InstitJte of Linguistics.)

10 PAGE 3

work in not more than four languages at anyone time. The trainingwas therefore set up as a aeriesof "cycles". Each cycle consisted of: 1. The initial training of the language officers in thebasic principles of constructing pre-primers, primers, and post-primers in theirown language.

2. The writing of suchmaterials in their own language.

3. The preliminary testingof these materials in two languages. classes in each of the

4. The revision of the materials in the light ofthe preliminary testing. 5. The publication of a limited trial edition of theseaaterials. 6. The training of teachers to use these materialsin the primary sckools.

7. Further revision aftertwo or three years'use in the schools

This whole process wasto becarried out in four Naturally, priority languages at a time. was given to theRole A languages. Four therefore chosen from this languages were group for Cycle 1: Bari, Dinka, Lotuho, andNdogo. The Ministry of Educationchose teachers who were designated and were to receive the language officers training and carry out thewriting. It was understood that a secondcycle of another four Role A languages would startas soon as progress with the first cycle made this possible.There would then bea third cycle which would completethe Role A languages and include languages. some of the Role B

Thoughthese plans had thefull supportoftheSouthern Government, there were Regional severe practical problems to beovercome. Government was destitute The Regional of resources. Many buildings had been during the Civil War. destroyed After much searching,a temporary solution was finally found. Some buildings atPalotaka,which were notin use, were made available. In June 1977 the first training began with personnelfrom the four Role A languages of Bari,Dinka, Lotuho, and Ndogo.

It soonbecame very evident that the facilities in Palotakawere totally inadequate and the location unsuitable.Palotaka lackbd all anenities and there were no supplies in the area. Everything had to be transported film Juba. More seriously, the location was completely isolated from theMinistry of Education. It was essential that the Institute of RegionalLanguages work closely with thoseresponsible for curriculum development andteacher training. Those activitieswere being carried out at Maridi, days journey away. a small town two There was no way activitiesat Palotaka and Maridi could becoordinated. The state of theroads (none werepaved), vehicles and fuel, the lack the lack of of any postal or telephonecoamunication, all made coamunication virtually impossible. It seemedessential for the Institute of Regional Languages to be relocated at Maridi. There itcould work alongside the otherinstitutes responsiblefor the education. development of primary school

Thefacilities at Maridi were already stretched to the Institute of Regional limit and the Languages could not be housed there without additional

11 PAGE

building. Although the Ministry was fullycommitted to the project, it did not have the resources necessary to construct the additional buildings. What limited funds it did have were morethan completely takenup byregular expenses such as staff salaries.

While the immediate pressure camefrom thelack of facilities where the proposed training could take place, therewere other aspects ofthe project which also needed financing. The original agreement between the Ministry of Education and SIL had stated that while SIL was respo5sible for the expenses of the overseas personnel, "the Ministry shall be responsible forthe expenses of local personnel and the cost of publication ofthe literacy and transfer aaterials". At a very early stage of the projectit becameapparent that although theMinistry would continue to pay thesalary of theteachers who were requested to participate in the project,there was no aoney for essential expenses such as the travel costs to enablethe teachers to go to the location wherethe training wasto be carriedout, nor wastheremoney for the addiitonal costs they incurred by theirbeing awayfrom theirhomes.The Ministry did not have funds either for essential equipment, forexample, there were no vehicles available for the supervision of the first test classes. Nor was there money to pay for publishing thematerials. Thiswas the situation when theRegionalMinstry turned toSIL and requested itsassistance. SIL isnot a foundation or a funding does not have agency and financial resources for suchexpenditures. Itwas apparent to all .concerned, however, that without outsidefunds the project could continue. not When the possibility of an approachbeing made to the United States Agency for International Development (AID) was raised, the Ministry welcomed this and requested SIL to go ahead as quickly as possiblein contacting AID and making a specific proposal.

In March 1978, work beganin Jubaon drafting aproposal in consultation with full the regional Ministry. Subsequently there were extensive discussions with the AID mission in Khartoum and the AID officein Washington. On 1 July 1979, AID awarded agrant totalling$1,400,000.Of this grant, 50.9% was allocated to the construction of the facilities at Maridi, 14%. to equipment,5.8% totravel, 5.4% to salaries, 2.5% to training, 2.4% to publication costs, 9% to administrative expenses,and 10% to theCenter for Applied Linguistics (CAL) for carryingout an evaluation of the project. All over Africa there has been agrowing interest in Childrenbeginning to learn toread andwrite in their own language ratherthan in the official language through which much of their education takes place. Given the degree of interestin this, it had seemed to mefor many years that itwould be valuableto make some evaluation of such an approachand if possible to compare it with the alternative, i.e., starting education by learning toread andwrite in theofficial language, asecondlanguage to the children concerned.Thus, the evaluation of theSudan project presented itselfas a good opportunity to investigatethis question empirically.

In a project ofthis nature, it isessential to build inregular evaluation.The educational materials need to be testedand evaluated. The Sudan project seemed to be an opportunityto combine the immediateneeds of a particular projectwith this broader need toevaluate such a projecton a wider basis, trying toassess the effectiveness of a multilingualapproach to

12 eires PAGE 6 1.1agaialMIEULMILEHM

1% =11151121

Oa eattegpeteeletashen Hoppe engaged inresearch in Sudan from March 011for to 1110. ale **teepresents data which he colleotedtogether with elsernattene MIdleemelenfly the authors. OM reeler* OSthe emits' eentest result et reeesentlatiene of the literacy projectdeveloped as a by Who and Coogan intheirsecond impact 44113M4ASS, WV MISIOStedthat furtherresearchbe done asong the Zande and Ober la alleiheapt fe determinethe reasons for the "...apparentlarge degree etveriMility hi %Mir attainmentof reeding..." (Csiko A Cowan, 41). Sea Femme@ were le 1682, p. be sought within thelarger culturalcontext, thelsdias aet ealpthe seheel but also la andoutside of the home. ihr Minim reseerela The idea was Unloosed later thatyear in a CAL-SIL meeting in heshiagesat 16L, whereit MS decidedthat 311. would support 'an te de OM ressereb. On April AO, 1912, in a letter toRichard teehealitir=savea math* statement of hls expectationsfor this research. Ihe alma masa eelass-sehoel context data will becollected by an Saloeupperted esarepelettatwho will bests working in the TOMO ones spereeteatniy Nasir and three mouths prior to lon's (Cowan)visit and esatthee We Weft atter lenosdeparturetor a total of approximately sis GOMM. . . Seheal metest data wouldomelet of information on esehestea/ wobbles operettaswithin the setting of the school(e.g., teseherswim* sad style,students' attendance, dented 60 li class sise, time skills).Nempeohool context data willconsist of imMemaklea sa sopesure to reeding and writing outsideof the SIMS' Settles, liElliavallebility amd me of readingmaterialsin the eltildratos hem, parental attitudestoward school and literacy,as well ether rasters Whigh appear to be having an influenceon the aeqpialtiM ise litmus skills lathe vernacular languages. SUM* OM NWer amd Samba were to be the focal point of thisresearch, Heppe telt Ahei letemellem fres tow other groups in SouthernSudan would :treader modersteadias et the literacyand educational climate in OW" as well m allowmore contrasts and comparisons of similarand dislar "sops. Osasequently,he smoldered adding the Kresh and Dinka to Me Mei, Abe ENOwere te be added because theyare an agricultural people lihe the bads, ha live isthe northern part of SouthernSudan and, like the Dm, were believed te spook more Arabic than most other SouthernSudanese. Ihe OMR isre theme Moms*of their cultural sisilarity pmemel to the Nuer as a amd bemuse they ware believed to bemore advanced Aham bhe hisr. Am fact, however, because of unforeseen time emehrelate molddo research on only me of thosetwo groups. On the basis et imismatlem resolvedafter arriving be awe in Sudan indicating the Kresh to arebleedsad less well kmom then the Dinka,Hoppe decided to add only Me Omit is ehe study. PAGE 7

2.2DESCRIPTION OF RESEARCH

2.2.1 RESEARCH DESIGN

A review of the previous documentationfrom the locallanguages literacy project was undertaken prior to formulating research questions. The major questions developed forthe survey encompassed the followingareas: importance of (a) the indigenous language inSouthern Sudan; (b) thedegreeof correlation between attitudes towardformal education and type of society, 1.g., pastoral vs. agrarian, urban vs. rural; (c) parental perceptions ofhow educational decisions are made and of how formal education might benefittheir children; (d) community interest inand preferences for written materials; (e) parentalinvolvement in the decisionsaffecting formal education; (f) the effect of formal education on thestability of rural communities; (g)factors affecting pupils' attendance andthe continuity of formaleducation; and (h) the extentto which teaching methodsin use are compatible with traditional educational patterns.

A questionnaire was developed to collect information on these topics. Fiye major categories of persons to be interviewed emerged: (a) teacher, (b) student, (c) parent, (d) townsperson and (e) government orschool official. Also, six major categories of inquirieswere developed: (a) demographics, (b) attitude toward languagedifferences, (c) attitude toward literacy, (d) literacy and the hone, (e) literacyandsocial structure, religionand politics and (f) hindrances to literacy. The questionnaire alsoincluded a section for background inforaationon each schoolvisited andon classroom observations.

The first draft of this SouthernSudan LiteracyProject Questionnaire was reviewed by eighteen different people. These includedSIL members, three Sudanese who were acquaintedwith theproject and the projectresearchers Cziko andCowan. The questionnaire wasthenmodifiedon thebasisof suggestionsreceived. Upon arrivingin Sudan, Hoppetried out the questionnaire with language officersand SIL staff before spendingsix days checking and revising it with speakersof severallanguages indifferent cultural settings. Additional information aboutthe schools and the cultural andsocial context in which they weresituated wasgathered throughmany informal discussions in towns, homes andwhile traveling andbydirect observation.

2.2.2DATA COLLECTION

The data were collected primarilythrough interviews with subjects whowere randomly selectedfrom lists by throwing dice. Teachers, studentsand officials were selected from lists providedby education officers, headmasters and local administrative officers. Parents chosenwere those of the children selected fromthe lists of students. Either parent, the one willingto be interviewed or the one at home whenthe visit occurred, was used. Townspeople

15 PAGE 8

were selected randomly as well. After a central place wasidentified in the town, usuallynear the marketplace, thedice werethrown toselect houses going away from the center in the four cardinal directions.This process was repeated with several throws of the diceuntil theoutskirts of town were readhed in all directions.

Heppe had initially hoped to conductinterviews on an informal basis, directing the conversation along lines that he was pursuing. However, because of thelength of the interviews, time constraints andthe involvement of a number of different people in the interviewing process, this methodproved to be impractical andso a more foraalized interviewingtechnique was used. Local authorities and the constituency with whom Heppe was working statedthat this was culturally acceptable. The interviewer attempted to besensitive to the interviewees and their responses, noting places where they seemed to be hesitant to answer, or where the inforaation seemed inconsistentwith the that already given. Interviews were conductedeither in the place of business the home of the interviewee. or

Three attempts were made to visiteach designated person or household. no one was in the office If or at home, the researchers proceeded tothe next personon thelistor the next house, respectively. Little troublewas encounteredin getting a random sample amongteachers, officials, or townspeople, but since school was not in session and many of the studentswere difficult to find or were out of town, the interviewersoften had to proceed to the next student on their lists inorder to get the number of studentsand parents needed.

During Hoppe's visit to Sudan from March through May of 1983he spent 14 days in Hasir (Muer), five days in Mayom (Muer) and surroundingareas, 23 days in !ambit) (Zande) and surroundingarea and nine days in Raga rest of his (Kresh). The time was spent traveling, writing up thedata collectedand getting inforaation from people in Juba and Maridi. Hoppe alsospent a week interviewing in Khartoum. Approximately 25%of the interviews were conducted in English without theuse of an interpreter.The other 75% were conducted in the local languages of thetowns on which Hoppewas working, or in Arabic, either when Arabic wasthe first language of theperson being interviewed or was preferred by the interviewee. Hoppe was assisted by the language officers of IRL, who were well-qualifiedinterpreters, as they were native speakers of the main language where the survey wasconducted and also spokeEnglish and Arabic fluently. Hoppe also had the help of additionalSIL people while in Yambio and Moir. His wife assisted him inNasir and Mayom andfor part of the time while in !ambit:).

The people in the communities where the interviewswereconducted were by and large very friendly and cooperative.Thiswas no doubt due in large part to the presence of the IRL languageofficers, who were well known and respected. The interviewers were refused threetimes in Yambio by people who suspected their motives, fearing that they were collecting information that mightbeused to incriminatepeoplewhogaveresponses whichmight be construed as critical of the government. Three people also refused to give interviewsin Nasirand there were eightinstances in which alternative choices had to be made because no onewas home at the houseselected for an interview, even though it was visitedthree different tires on differentdays at different times.Sixty-one interviews were conductedin /Wirt 28 in Mayon

16 PAGE 9

and Urach, a small settlement five miles outside ofMayon and in the village of Mankien 25 miles away. Eight interviews were conductedalong the road from Urach to Menkien. Eighty-six interviewswereconducted in theYambio area, including Yanbio proper, Maingbangaru, a village seven milesnorthwestof Yambio, Nzara, a villageseven miles distant from Maingbangaru, the Susa located three miles village of northwest of Nam' and Singbi,a village four miles southeast of Nzara. A total of 46 interviewswascollected in the Raga and the mall town of town of Argabanga, two miles northwest of it.In all, 227 people were interviewed including 97 Nuer, 86 Zande and 44 Kreshspeakers.

2.3 RESULTS

2.3.1 BACKGROUND DATA

2.3.1.1Nasir

Nasir isa Nuer-speaking town of about6000 locatedon theSobat River about 150 miles southeast of Milakal and 20 miles fromthe Ethiopian border. The townis spread out along the southernbank of theriver and is approximately two miles long and one-half milewide at its widest point.The airstrip used to be on the north end of town perpendicular to theriver, but a large residential area has grown up on the north side of the airstripover the past five years. Just south of the airstrip, on the bank of the river is the 'A! Commissioner's house. Alongthe river heading east,interspersed with residences,are the police headquarters, market, Presbyterianchurch, hospital, military housing and the militarycantonment, respectively. The prison is located a block behind the police headquarters. The town also has a veterinary center, sports field, Lutheran WorldService Center, blacksmith, mosque, located next to the market area, police club, Sudan Socialist Union Centerand assorted administration buildings. In additionto theold marketplace, run mainly by Arabs, a new market area was started severalyears ago across from thesports field.Merchants in thenew market are nainly Nuers selling local items.

Nasir is linked with Malelal by a secondary road, which ispassable only during dryseason. During the rainy season steamboats come up the Sobat River to Nasir bringing cargoandpassengers. Thereare alsooccasional flights from Malakal to Nasir.

The school buildings mentioned in the third impact evaluationwere still in use but in very poor repair. The school consisted of threebuildings, all of which had mud ormud and mortar walls supporting tinroofs. Thefirst of these, a "U"shaped complex, housed classroom, theheadmaster's room,a storage mole and a teacher's room.. The other two buildingswere one-room structures approximately 80 by20 feet. Sitting on the dirtfloor or stacks of bricks they brought With them, as many as 120children occupied these one-room schoolhouses.

17 PAGE 10

At the time of Hoppe's visit,he was told that 150 P1 children one of had occupied the classrooms about 20 feetwide and40 feet inlength and students were taught 98 P2 in another room approximately15 by 30 feet previous year.The during the crowdedconditions were caused bythe influx of new students due to activities of southern resistance forces, theAnya Nya, in the surrounding towns and countryside. An increasein students was expected the coming school year. in

The interviewers' intention hadbeen not only to survey thetown but also several of thecattle camps, but this was not possible, however, due to the activities of the Anya Ilya. The policechief, concerned for would not allow the interviewers their safety, to venturebeyond the perimeters of the proper. During town the interviewers' stayin Nasir, the Anya Nya aade several raids around the town, taking numerous cattle and shooting upa local vehicle. Acurfew from 6 p.m. to 6 a.m.was placedon the town; and on occasions, the interviewers several were ordered to return to theirhouse because of Anya Nya activitiesnearthe town. They were fortunate, however, to to be able interview people who hadrecently movedto Nasirfrom the surrounding countryside.

2.3.1.2 Mayom

Mayom is a rather new small Muer settlement of about350 inhabitants that has grown up around thenew government center constructed in 1982. out in a long, narrow band It is laid about one milelong by three blocks wide andhas a market area with abouta dozen shops, a police station and next to the government offices an airstrip located and the mission area.There was no school, but it was hoped that one would open soon. Mayom is connected by car tracks with Mankein 23 miles to the south. About 70 miles to the easton a secondary road lies Bentiu, where a school with grades P1.through P6and a junior secondary boarding school are located.

The security situationIn theMayom area was somewhat first, the "A" precarious. At Commissioner did.not want to allow the interviewersto leave Mayon proper, explaining that the Anya Nya had stolen 90head of cattle from Mayom judge the a day after the interviewersarrived and that the the process of making army was in a military sweep through the area.While in Mayon, the interviewers visitedthe village of Brach, a small people five miles settlement of about 75 from Mayon, where they gatheredbackground information and conducted six interviews. They also visited the villageof Mankein, which has a population of about 300 and were also able to do sixmore interviews. thereturn trip On from Mankein, theinterviewers dideight 1040-15 minute (shortened) interviews, stopping at every sixteenthhouse (chosen along the car track. by dice)

18 PAGE 11

2.3.1.3Yambio

Although the main ethnicgroup located in and around Yambiois the Zande, meabers from at least 19 other tribeslying therewere talliedduring the research. Yambio is a town of about 10,000 people, ipproximately 20 miles from the border of the Central AfricanRepublic inthe extreme louthwest corner of western Equatoria. It is the capital of Western Equatoriaand the provincialand district education headquarters. There are numerousother government departaents located inYambioincluding agriculture,health, wildlife and tourism, communications, forestry, socialwelfare, police, district prison, coamerce, veterinary medicine, Sudan Socialist Union and labor.

In addition to a number of primary schools, twojuniorsecondary schools and a senior secondary school, thereis also an agricultural training and experimental center located in Yambio. An airstripis located next to the agricultural center. Yambio hasseveral grist ills and a largemarket center. A new mosque was under construction next to themarket. The church missionarysociety of the Anglican Church opened its firststation in Zandeland in Yambio in 1913 and the Anglican bishop ofwestern Equatoria now resides in Yaablo where they havea large church and primary schoolon the southwest side of town. There is also a Catholic churchlocated on the west side of town with a resident priest. The Catholics were planningto open a school soon.

In Yambio and the surroundingarea, the political and religious atmosphere appeared, more relaxed. English was spoken more in Yambio thanin any of the other areas visited except Juba, theformer regional capital. Also, the local Zande speakersexhibited much pride over theuse of Zandein all areas of life, butspoke Arabic when necessary. Theydemonstratedadefinite antagonismtoward the fact thatArabic wasdominantin theschool anda majority of the people expresseda desire to use the vernacular in the schools exclusively for at least the firsttwo years of schooling. People repeatedly asked about getting sore material in the vernacular language .and especially expressed a desire to have theirownhistory andfolklore in theZande language as part of the school curriculum.

2.3.1.4 Bag!

Raga, located in theprovince of Bahr el Ghazal about160 miles northwest of the provincial capital and 50milesfrom the border of the Central African Republic, ispopulated by about 4000 inhabitants. It is connected by road with Wau and has an airstrip.The townseeas to have two centers, the market and commercial area forming one andthe government buildings, including the districtpolice station, educationbuilding, prison andhospitalforming another. The Catholic church and a mosque arealso located inthis latter area.

The main vernacular languagespoken in the areais &esti, although the interviewersrecorded 11 vernaculars, plus Arabicand English. Other

19 PAGE 12

important vernacular languages in the area are Feroge andBanda. Arabic is lore widely spoken in Raga than any of theareas Hoppe visited, except Juba. There was a greater sense of fear and tensionin Raga than in the other areas visited, as evidenced by the fee':that people were sore secretiveabout their answers to the interview questions and showeda narked distrust of the central government. People were especially fearful of thegovernment and its policiesconcerning education, security,andgeneral administration. The interviewer was told that a large number of Nreshwere killed by the Arabs during the war, which has fed theantagonism of tribal peoplesagainst what theyfeel is an attempt by thecentralgovernment topressure theminto Arabization and Islamization.

2.3.2 QUESTIONNAIRE DATA

2.3.2.1Language Use and Attitudes

The survey recorded over 50 different vernacular languages, in addition to Arabic, English and several otherlanguagessuch as French, Bango, Lingal and Aaharic spoken in surrounding countriesand in Southern Sudan. percent of those Ninety-three interviewed felt mostcomfortable speakingtheirown vernacular languages, but because of theheterogeneity of languageusage in SouthernSudan, theyfound itnecessary tospeak atleast one or more additional languages. At least 83% of the people spoke two languages and 56% spoke three or more languages.

If language context is viewedas homogeneous when a single language exists in a given context and heterogeneous when a number of languagesare presented, a continuum may be constructed in which language context in the home falls at the homogeneous end of the scale andlanguage context in institutionssuch as the military fall at the heterogeneousend. In between would fall the market, recreation, town meetings and differenttypes of work other than the military. This continuum is depicted inTable 2.1.

The expectations for language use vary according to context. For example, in the home csingle languageis normally used, with an additional used language(s) only infrequently. Inthemilitary, on the other hand,since a heterogeneous situation exists, a lingua franca isused, inmost instances Arabic. In the military heterogeneity existsnot so much inlanguage usage but inthe nuaber of languages present, which forcesthe use ofthe lingua franca.Because the official second languagein Southern Sudanis English, the expectation for languageusage in most officesis English,although a considerable mount of Arabic and vernacular isalso used. In other types of heterogeneous language situations,such asrecreation, market, or different types of public meetings, the language is usually selectedon the basis of a common language. If a person's mother tongue isa vernacular language, but he also speaks Arabic, English, and/or severalother vernacularlanguages, he will usually speak the language of theindividual to whom he is talking. For example, in the market, if a Zandespeaker is speaking toanother Zande, he Table 2.1

Percent of RespondentsReporting to Speak Various Languages InVarious Contexts

Contexts

Home Public Language Government (Homogeneity) Recreation Military Market Work Meetings Meetings (Heterogeneity) OMMIMM11.MIIIPPIPMINMPIIMM.I.MIMMIMMIIMMMINIMUMMINEW

Vernacular 91% 451 371 47% 45% 27% 7% Arabic 3% 25% 301 15% 161 19% 33% English 1.5% 5% 5% 7% Arabic and English 6% 2% 111 13% 22% 53% Arabic and Vernacular 3% 51 16% 4% 31 5% Arabic, English and Vernacular 4% 1.5% 6% MI 7% Arable and Other Vernacular 6% 13% 3.5% 3% 3, Other 3% 7.5% 0.5% 8.5% 7.5% 17%

Note: Data collected Inresponse to the question, "What language doyou speak In X context?"

21 22 PAGE 13

will useZande. If a Zande speaker is speakingto a person whospeaks a vernacular he hiaself does notspeak, orArabic, theywill WO Arabic, or another cossonlanguage which they can bothuse and understand. In public meetings, interpreters areused to communicate withthosewhodo not understand the primary languagebeing used. Mat Southern Sudanese recognize the necessity of havinga lingua franca to communicate with one anotherintertribally. The problem, of deciding whidh language should course, comes in be used. Some of the larger tribal groups would preferthat their language become thelingua franca. Other Sudanese prefer Arabic and some prefer English. Table 2.2 provides a breakdownof the responses received in eacharea of the survey to the question: is considered Which language moat prestigious? Note that 61percent of chose the local vernacular, those interviewed while only 16% chose Arabic and 13%chose English. (The other 10%, except for 2% who chaseArabic and Englishtogether, Chose somecombination of two languages whichincluded a vernacularlanguage.) Eighty percent of the Muerand 51% ofthe Zande said the most prestigious. vernacular was the Onlyin Raga did people selectArabic as being the most over the vernacular presitigious. Even after a hundred andthirty years of Arabic and Englishusage in the Southern Sudan, mostpeoplestill consider their mother tongueto be most prestigious and very important to them. The vernaculars embody a personaland cultural identity which fear will be swallowed the people greatly up inthecompulsory use of a lingua franca. Therefore, although 55percent felt that it was advantageous to read and write for them to learn Arabic, 95% felt that itwas also advantageous fOr them to learn to read and write theirown vernacular languages.

Less positive attitudeswere disPovered with respect to the in the South. use of Arabic Sixty-nine percent o: thoseinterviewed said they felt that sain language the being used in theprimary sehools, i.e., Arabic, education of the children. hindered the Also, 67% ofthe teathers said that theybelieved the children would leafn',otter, especiallyin the first several the vernacular. grades, in Eighteen pment of the teacherssuggested two of the languages while a combination of only 15% said that students learn best In the same vein, 44% of in Arabic. the interviewees felt that Arabicwas a direct threat to their cultural heritage.

2.3.2.2 Urban-Rural and Ararian-Pastoral Differences

The second and third impact evaluationsreports indicEted a superiority the performance of the Zande in children over the Nuer Children. Since the Zande are an agriculturally dominated society, whilethe Nuer are essentiallya pastoral tribe, one possible hypothesiswasthat thedifferencesin educational attainment might be a result of attitudestoward educationin thesedifferent societies. A second equally obvioushypothesis was that attitudes toward education might differ markedly dependingon the location -- rural or urban-- of the student population.

Of those interviewed, 59%came from rural areas and 41% In this study, 'urban' from urban areas. was arbitrarily defined as townslargerthan 4000 and 'rural' as anything smaller. The data revealed very littledifference between

23 rmuc. s a

Table 2.2 Most Prestigious Languagea

Language

Location Arabic English Vernacular Other

Nasir-Mayom (Nuer) 6% (9) 9%(8) 80%(70) 5%(4)

Yambio (Uncle) 12% (10) 21%(19) 51%(43) 16%(14)

Raga (Kresh) 45% (20) 5%(2) 43%(19) 7%(3) a Percent (frequency) of respondents selectinga given language in response to the question "Which language is most prestigious?"

24 PAGE 14

urban and rural felt needs for education. Ninety-eight percentof the urban people said that they feel it is goodfor everyone to learn to readand write and of the rural people, 94% said that it is good for everyone toread and write. In the same light,among the urban people, 96% said they believe good for the community it is to have a school and 97% of therural people said the same.

Most people stated that the community leadersare interested in education for their communities.Of the urbanpopulation, 61% of the interviewees that thoy believe the leaders said are interested and of the rural population,69% of the people said they believetheleaders are interested in the educational concerns of thecommunity. Eighty-nine percent ofthe urban and 94%of the rural people stated that theybelieve the community encouraged schooling. Examining the above data along thelinesof agricultural (Zande)and pastoral (Muer) also shows no marked difference between the twogroups. Both groups show a very high interest in education and literacy, ascan be seen by Table 2.3.

It was assumed that a difference might exist between the urban and rural pastoral societies with regard to worldview andbasic values. Time constraints aade it impossible toresearch indepth the world view and values of the Nuer, but in regard to value for and interestin education, Table 2.3 also shows no great difference between urban and ruralMueremerged when the Muer of Masir are compared with the Nuer of Mayon and surroundingregion.

2.3.2.3 Reading Interests

By far the greatest interest voiced in the differentgroups interviewed was for printed materials about tribal.histories, folkstoriesand folklore (see Table 2.4). There were also a number of requestsfor such topics as hygiene, health, disease, cattle,:farming, religious books,and Bible stories. The main request, however, wassimply fOr vernacularmaterial. Minety-seven percent ofthe people, when asked if they wanted tribalhistories written down, answered in the affirmative.

2.3.2.4Availability and Control of Education

Seventy-four percent of the people interviewed said there wereno adult literacy classesin the coamunity. Most expressed a desire forsuch adult literacyclasses. The peoplewho were aware of an adultliteracy functioning in class their community werein most cases referring tothe classes held'in the local church. None of the government adultliteracy classes were functioning at the time.and had not been functioning forsome tine. Questions such as, "Who decides if there will be a school?"or "Who decides what willbe taughtin the schools?" were answered bya largevariety of statements, a number of them contradictory.One third of thepeople gave no

ArALCV,.. PAGE 14 a

Table 2.3 Attitude Toward Education and Literacya

Language Location Question Nuer Zande Kresh Nasir Mayom

Community encourages education? 97%, .M11. 98 88 INM

Good for community to havea school? 99$ 97$ 914ii 98$ 1 00$

Good for everyone to learn to read and write? 98 1 00$ 90$ 96$ 87$ a Percent of respondents responding in affirmativeto each question.

,

26 liwboi eimiNe eseinmint

Ivy. so Om, ftesooliro. 116111111111111=11.111111111111111111111111..

%so ye ei am) $a

amp* 3. ampownsmowerame

WOO SW UM

OWIII04 411MO Md.%

41111111411IPS 11,11.111100 ell

OM SIMI OS *000SIM 3.

SOP' %War "Nib 411111111.11.11P Os saiseim *Ow mewl 0 MOM WOO 406Criall)

17 gth.L: PAGE 15

smear to these questions and only onethird answered all fivequestions (tmentitthe theselaterviewed were either townofficialsor tuaers). lhis seems school to Wilma that the interviewees,for the most part, 0110 111* suers st hen the sohool la run and thatthey are dependenton ettivals or livormht mithority Inorder to have a school in their ia location. very real sass, this seems to be true. In all the areas visited,except plirerit=irektek hod me sant, the peopleindicatedthere was a le ammoisties, butin none ot thecommunities was this assesistia sassily tastismimg. The fact that people still viewthe school stersigh. salty amounts ferone of the aain reasons tor thelack of press rests latereeties between the community and the school and betweenthe samaity ad the govermat,astar as the educational thus process is concerned. Mied *ether the seek helpforthe school apart from 111e levereseekt 111 st the answered "Yes" and another18% said "Yes", as laces it is im supers nith the government or other outsidesources. OW MU sold thmy did met thinkhelp should be sought trom theoutside. them asksd Mall the government closely oontrol theschools?", 56% imdisstod tha it weld lie good itthe community had some cooperativerole in the aslial. Omt it MSobserved that the community tooklittle responsibility ter the ahem'. this mes seem In the poor attendance recordsand facilities sum entsim ostremsly peas repair. Parents could haveimprovedthe hollemps, provided hashes, tood,pencils and notebooks, but theyfelt it was the govaramemtes reopmestbility to do so. even in the tew self-help schools the MO thtmg that mows to make them different free thegovernaent schools IM:mjeeplie im. She immunity ere Gownedenough toinitiate an programWfWiliam sohoolhome. Otherwise, the teachers, malarieS, meterials(esmak as are available toany government school) and serrismime admildstratiouare all provided by the government.As Michael O'OrieR, the adainletrater ot theLutheran Morld Service project maid in Hasir, ells must isok at the presentselt-help concept in Southern Sudannot as the flashed predmet but as a system in transition which is heading in the riget dirsaism. This divided perception of material provision and 1111114111a1111 SWIMindicates that the people viewgovernment school support tea dmal perepective. On the one hand, they expectthe governaent to =the phyeieal ascesalties tor thechildren's school while on the other hey meet a part la the decisions otwhat willbetaught in the elasereem.

Owe etthe goals ot this mem*was to deteraine whetherthe current teaohing methods used in the schoolsare indigenousto thelocal area or uhether they 001111down through the educational tastes. Duringthe time this reeler* weebeing outdated in Southern Sudan,none of the schools were open. This nodeeetmal observation of the teachingmethods impossible.However, dissensions with a member ot the teachersdid shed some light on the subject. At present,meetetthe primary school classesin Southern Sudan are using methods Wish depend very heavilyupon the use of the blackboard as a visual aid, rate memossinstion, amd specificreproduction of what the teacher gives in OASES, SSmeted in the tiret and second iapact evaluations. In general, the

28 4k! LA!! PAGE 16

student is not taught to think for himself, pursuingproblems or situations, but onlyto reproduce that which the teacher gives.Garvey-Williams in his UNESCO report identified the problemsresulting from the teaching methods:

Learning is made to depend heavily on memorization and teaching leanson oral instruction, making littleor no use of visual aids except the blackboard andproviding verylittletraining in reasoning and application. (1976, p. 39).

It would appear that traditionalmethods of education cannot be associated with the concept of "formal" education,since the only type of training that the children habitually receiveis of a more practical natureand is usually taught on an individual level withina family context. Among the Zande, when a young boy is brought to a prince's compoundfor training, he was assigned to an older boy or man responsible for instructingthe young boy in all aspects of the prince's court and in methodsof warfare as well asother cultural ways.

The training was in two distinctphases. At the age of perhaps ten, a boy would become assistant toa young man who was already a full member of a company. The boy would look after the physicalneeds of the . youth (EVans-Pritchard, 1971,p. 199), fetching water, tending the fire, bringing the foodwhichthe womencooked, bearing his shieldwhen traveling, looking after his few possessions....

In return for these duties, theyoung man would beresponsible for the boy's early education. At the age orseventeen or eighteen,the boy would become a fUll member ofthe company....The patterns of Zande life would suggest thatthe knowledge and skills which would then be passed own tothe young menby their commander would include the following: religion, farming, hunting and fishing, building, language, history, music and dancing. (Atkins, n.d., p. 9)

It is impossible to determine withany certainty how thecurrent methods usedin schools evolved. They are nodoubtpartly alegacy of colonial educationalsystem, partly a reflection of currentteaching patterns and partly a reflection of the Arabicsystemof education. What is noteworthy is that the methodology presently inuse does lictle to encourage the development of individual initiative for learning tolearn and that Hoppe interviews with educators substantiates thepicture presented in the secondand third impact evaluations.

2.3.2.6Iaportance Of Education

Because of the cultural and religiousimplications of Arabic in the South, attitudestoward the importance of education werefOund to be somewhatof a - paradox. Although100% ofthe Southern Sudanese interviewedexpressed the belief that it is good, if notnecessary, for the people of the South to have education72% expressed doubts concerningthe educational process, primarily due to wh:tthey see as the Arabization. of the South.. Inother words, the

29 PAGE 17

peoplefelt that they need an education tobe able todirect theirown destiny, but they feltthat the typeof educationthey werereceiving was destroying their cultural identity. Because of this conflict,some parents reportedly choose not to send their childrento school.

Similarly, 95% of the people said.thatit was good for everyone, adultsas well as children, tolearn to readand write, though sone did exclude the .elderly and a few 'excluded girls. Concerning thechoice of language with respect to reading and_writing, 95% said they would like to read and writein the vernacular and 88%in English while only 44% reportedthey would like to read and write Arabic.Overall, 69% of the people responded thatthey felt Arabic was a hindrance tothe literacy program. Further evidence of these mixed feelings towardArabic ahd the North was found in responses indicating that while 87% ofthe people feltthat the central governmentin Khartoum encouraged 'literacyinArabic, only 33%feltthe government encouraged literacy in the vernacular.

2.3.2.7Consequences of Education

In addition to assessing communityfeelings on the imi;ortance of education, this survey sought.to determine whatSoutherwSudanese thought education would bring them in teras.of material benefits. The main reasons given for getting an educationvere -the-sake ones discovered in thesecond impact evaluation, to get a .job,:lo become a big.:person,to help the country and to helpone's people. All:of these interests basically funnelinto one and the same goal-- having some -type of job ..with.the government, .because the government is the. primary 'employer andbecause working for.the governmentimpliesposition, prestige and power. This interestis in no waycurtailed by the fact that government wages are not very high.Getting a job with the government is the normal expedted outcone of the educationalprocess and is in fact considered a worthy-, aspirationvtherefore, this seeasto havein noway decreasedthe 'Interesk:In education on the part of thechildren or parents.

The resUlts orthe survey indicate thatthe acquisition of formal education produces-an.exOdus from-rural areas. Sixty-four percent of thepeople said that an. educated person doesleave home. In actUal practicethis normally means that. In order to .get a job.theperson must be assigned to a job by the government and this'.,could be in anyarea of SOuthern Sudan. If he wants to worko.the'employee must -take the assignmentwherever it is. Since there are few goirernment :positionsopen in thesmaller communities, thereare usually not many edUcated persons available for children tomodel after, although there usuallTare a-few inthe communities. Inmost ofthe placesthe Interviewers_ visited, 'thereWere several teachers andtown officials. Nowever,,,the.:.children are continually encouragedby the teachers, by their parents and bylhe tOwn officials to be educatedao that theymight become people who can helvthe country and helptheir community. Accordingto the interviOai, teaChers averagedtwo yearsin a given geographicallocation. This chntinual teacher movement- caused alack ofcontinuity inthe school teaching 'staff and hampered parent -teacerrelationships.

30 PAGE 18

2.3.2.8 School Attendance

The data from the interviews indicatedthat attendance is by and large low to expect any consistent gains too fromone year to another.Parents are not always aware of how often childrenshould or do attendschool. the In answer to question, How many monthsshould yourchildgo to school each year? Twenty-four percent of those answering did notknow. Of those whogave an opinion, 66.9% said nine months. When asked: How many months did your child go to school last year? Twenty-nine percentdid not know how many monthsthe children went to-school in 1982. Sixty-ninepercent (of those who knew)said that their child went to school six months or less in 1982. This indicates that over two-thirds of the people who seemedto know believe the children only went to schOol two-thirds of the specified tine. An interesting point of interpretation here from the perspectiveof many of the interviewees if the was that childrewwent to schoolfor just a few days.out ofany given month, that was considered a month of school. -Theydid not addup thedays and compare those.with a 20day school month. Instead, if a child went to school two or three days in any givenmonththat was considered a month of resulting in high.estimates school, in the number of months thechildren attended school. A chedk With the education administrators in !ambit) Raga and Nasir indicated that the Children were supposed to attend school 200 to 210days out of each year during.the months ofApril through December.

Most parents professed .to sendtheir children to school but a looltat every school day, theattendance records indicatedotherwise.The attendance records checked for, a nine mOnth period *the 1982 schoolyear showed that about 50% of the, children enrolled went to-achool onlyone-third of the days school was .in session. Approximately one quarter of thechildren attended between one -, and two-thirdsof the. sessions, the other quarterattended two-thirds or,more Of. therdayithe school was in session. It should be noted that moat attendance was.not consecutive,but broken up into periods ofa few days to -a.few...Weeeks scattered-over the nine monthperiod. Inthe Yambio area, ,the maxima 'nuaber of daysonesChool wasopen in nine months was seventy-six dayi.. AnOther showedsixty-five days, athird 52 days and a fourth thirtp.eight days. .In the'fuer area the maximum was eighty-seven days in nine months. ,The Raga school record showed eighty-one days insession for a nine ..month period. 'Of thesix primary -school attendancerecords (1982) checked,,the. schoollvaveraged .67days of attendance or33% of the200 days required.sehool attendance. Forty-three (87 days) was the highestnumber of days attended and:19% (30 daysYwasthelowest. This means an average of 25% ofthe children :attended school-45-67 days out of the nine months; 25%. attended 22-45 days Out Of the nine months and the remaining 50%attended less than 22 days in the 1982 schoolyear.

Teacher strikes, 440ot-shortages and subsistence activities accountedfor much of thelost time. Rost students who attendedschool at allin a given year, even if-only for a few weeks, werepushed on to the next class when the following school year began. "Promotion within each level ofeducation primary; junior secondary, seniorsecondary is automatic fromone grade to the s(Oarvey-Wiliiams. 1976:90).Even under idealconditions, children attending school soinfrequently would find it impossibleto learn and build knowledge froa one lesson to thenext.

31 PAGE 19

Another problem which surfaced repeatedly in the interviewsconcerning attendance was that over 50% of the P1-3students in all three areas said they went to school hungry.School began at 7:30or 8 a.m. and was dismissed from 9 to 10 a... to allow the childrento return home for breakfast, but some students did not return to school afterbreakfast. This was especially true of those Children who had a long distance (twoto four miles) totravel to school, but also included many who did notget fedduring breakfast hour because the parents were not homeor no food was prepared.

One thing that became quite clear in the interviews with theparents was their tack of understanding ofthenecessity fordaily sequentialschool attendanceto facilitate the learning process. The conceptof oneday's lesson building upon another did not seemto have culturalrelevance. The parents seemed to think that the childwould atleast learn half ofthe information involved if he went half of the time, whether they went three months Weight,' Or whether they attended sporadically. A further complication became apparent whenwecheckedthe parents° answers in the interview Aigainst -the attendance record at school and againstwhat the children ihterviewed-were .saying.Many parents said their children attended school most of: the time. The attendancerecord showed that none of the childremattended 100% of thetine school was in sesaion, althoughthere were a.few _who had near perfect attendancerecords. Almost 50% of the children attended:one4hird or less of the time thesehool was in session. (This was less than one eighth.the sPecified tise.) Since school was only in sessionan average of 67 days.for P1 and P2 pupils, 50% of them received a scant 22days of schooling for,an entire.schoolyear.

When asked about_this discrepancy, theparents wou'A either claim ignIrance about.the matter, oi say that sometimesitwas necessary for them to send the

childrento take. Care . of thecattle, protect the cultivated fields,run errand; or watch.a younger child.,Onthe other hand, when the childrenwere asked,"theyindicated that they sometimesstayed awayfrom schoolwithout their.parents' periission or theywere requiredby their parents to attend to various family considerations. Butthe main reason given forthe children missing-school was thatthe teachers were on strikeorabsent for sone other reason.: When the parents were asked if they understoodwhyit was necessary for the children to attend schoolevery day, most of them indicated that they did not, except that the teacher saidit was necessary,or theygave some alternate ansOerabout the childgetting a good education so thathe could have a better job; none gave anindication that they understand the mechanics ofthe educational process and the necessityof building one lessonupon another.

2.4 DISCUSSION

Theresults of this survey confirm theobservations regardingteaching practices clrrentlyused inthe schools which are recorded in considerable detallsin theSecond and third impact evaluations. Teachers admitted freely thatthe methodology stresses activities which de-emphasize productive learning. Hence this finding lends support to therecommendations for changes in methodology and materials ifany but the, saallest percentage of children

32 PAGE 20

(I.e., the most highly self-motivated)attending schoolare toattain functional literacy in theirindigenous languages. In addition to this anticipatedresult, the anthropological survey produceda sorecomplete picture of the sociological variableswhich Influencethe impactof formal education in rural schools. Due tothe the restricted regionalcoverage of the survey, however, caution needs tobe exercised in generalizing the results to the entire popul:tion of Southern Sudan.

One of the more Interesting and totallyunexpected results of the survey was the absence of a significantdifference in attitude towardeducation between urban and rural and more'particularly,between agrarian and pastoral societies. The data makeit appear that education is valued by all segments of society in Southern Sndan,even though parents,show an extremely vague idea of .what. benefitsare toaccruefrom -sendingtheir childrento school. Similarly, ,Although, there, isclearlybroadsupportfor literacy in the IndigenoUs languages,'-just how reading and writingmightbe ofismediate practical use in daily life does notseem to be evident to many adults,,as is evidenced by the preferences expressed inTable 2.4. A far greater percentage of the people interviewed see literacyas a vehicle:primarily for maintaining tribal customs and historythan as a means of obtaining knowledgewhich will Improve their standard of living.

These two findings4uggest that steps need tobe taken to sharpen community perceptions Of the practical benefits thatcan accrue froa literacy. It does not seem likely that the initial enthusiasmfOr the literacy project can be sustained over. many years unless all membersof the community can see it's value in some aspect of their daily existence. One of thefirst steps that should be taken isto discover what kinds of traditionalactivities might benefit from the introduction,of readingand writing. The results of Hoppe's study have projected some areas that might be explored in greater detail with a view to establishing the value of reading and writing inrural life.

The survey data suggest that parents wouldlike a greater role in aaking decisions about school but,at the same time, view themaintenance of formal education as thepole responsibility of thegovernment.The legacy of the colonial administration, which establishedthe concept of formal education as the responsibility of an "outside" or "foreign"group, has been transferred to the 'national government. 'Comments.; generated in thecourse of interviews indicated that an-external organisation shouldLave the responsibility for the school* 'just tathey have had inthe past, since the community neither hasthe know-how or resources to run'thes.. Theobvious implication here is that the Ministry should be aware of thisview and that certaintypes of programs designed tO shift some of the burden ofschool maintenance, e.g., "self-help" programs will meet with little enthusiasm.

In addition to thebasicproblem of large numbersof poorly trained teachers, which obvioUsly affects the, calibreof instruction pupils received, the sUrvey -confireed something thatt.he educators whoparticipated inthe impact.evaluations frequently complained of--sporadic attendance by teachers and pupils and frequentinterruptions throughout- the year (often due to strikes), thereby, curtailing the totalamount of instruction plannedfor a year as well as the continuity necessary for learning. Teacher absences are due not only to the fact that payIS so low that primary school teachersfind it necessary to supplement their salaries with gardening,building their own PAGE 21

houses and involvement in other subsistenceactivities, but also because they genuinely want to farm or raise cattle.Many of the teachers interviewed said that they viewed these activitiesas a regular part of their life and heritage andcould not imagine doing without then. As a resultof this attitude absences froa their western-style job, i.e., teaching, were inevitable and in their opinion quite justified.

The survey revealed a number ofattitudinal factors which combinedto seriously interrupt and impede the continuityof formal education in Southern Sudan, in particular in the'ruralareas.These attitudinal barriers result fromclash betweentraditionallife style values,to which the Southern Sudanese cling- tenaciously andthe demandsimposed by a westernstyle educational system. This clash it manifested in the behavior of educators, parents, and children.One example isthe parents' attitude toward education and their recOgnition of thOieed to recruitthe largestpossible work force to sustain their subsistence economt- life-style. Although parents overwheliingly endorsed the concept:of formal education, they also recognized how Useful their children could be 'helping in subsistence pursuits. Consequently, many parents subscribed togovernmental educational policies and gave them 'llp-servicesuntildemandswere placedupon them, e.g., providing what is necessary for:- the:child togo toschool: clothes, regular meals, schOol:itterials :and actuslly losing thechild fromthe family workforde. Daily nedeapity.then began to dominateand children missed school frequently since many parents were-not ableor, attimes, notwilling, towithstand further economic.-deprivation. Th factthat the schools in their area had poorly ,trained teachers.and no suppliesand generally did not live up to their ideal'ef what aZchool. should'be did littleto encourage them to sacrifice additional..workeribr.sending, theirchildren to school. This problemis compounded by another-finding of. thesurvey - - the fact thatparents do not comprehend,thnnecetsity;Of continuous attendance.Thus, those parents who do send their children to,school often believethat extremely sporadic attendance is acceptable and represents an adequatesacrifice on their part.

Poor attendance Was obviously .in part due tothe same reasons whichcause children throughout-the worldto shun !Orvaleducation, i.e., lack of interesting activities orchallengeprovided by theteacher, learning mtd:.:.Dut the survey revealed anotherimportant cause:a clash between the traditionalsubsistenceschedule' andthe western education schedule introduced during the colonialera. As many as 50% of these children did not return after the morning breakfast breaks becausetheylived too far away and their parents were occupied duringthe school recess and could not prepare food for them.' They therefore had to seek foodin their compound or in their parents' fields.

Some of the attitudinal conflicts mentionedherecanbe treatedby government intervention, but othersmay bemore difficult to deal with.Many Sudanese educators argue that the reinstitutionof the boarding school system at the 'primary school level would do much to reduce truancyresulting from children having to. leave schoolto find a source of food. Setter and regularly paid salaries would doubtlessmake teaching more attractive and reduce teachers' dependence on outsidesources of income. It should also be noted that regional governments have attemptedto accommodate pressures which sten froa traditonal culturalpatterns throughthe institutionof official holidays, such as "planting day" in WesternEquatoria province. Whether a PAGE 22

carefully planned and well executed campaign to educateparents about the necessity of insuring that their children attend school "regularly" (aconcept which obviously needs to beredefined formany adults) wouldbe successful depends inlarge part on demonstrating the practical value of educationand the activities of the schools to provide it. The survey showed that at the present time neither of these conditions are fulfilled in the oyesof most rural Sudanese. Educationis presently sustained throughvague promises and hopes that good things may come in tbefuture.

An issue that surfaced time and again in the interviews was theproblem of discipline in the schools and at home. Lack of discipline in the schoolswas usually blased on thenorthern Arabic influence in theeducational system. Many tribes-people said, "The Northerners are using theschoolsto steal our children fromus." This was heard repeatedly, notonlythroughthe interviews, but also while travelingor visiting with people. Often as soon as people heard that an "educational survey"was beingconducted, this topic would come up. Parents saidthe lack of disciplinein the schools and the encouragement of students by northernSudanese Arab orArab-enculturated teachers to defy the authority structure in thehome was at thecore of the problem. Southern parents and townspeople said thatsince the Arabs could not defeat them in the civil war, they were going to do it through controlof the educational system and the economy.Thisconflict spawns a great deal Ibf latent'hostility and suspicion between parentsin certain areassuch as Bahr El Ghazal province and those involved inthe educational system and itmay be one of the primary reasons why many parentsregretted the absence of some organization such as PTA where they could express their views about whatgoes on in the school.

A surprisingly large number of people interviewedexpressed the view that the major problem with education in SouthernSudan is political. They felt that aany teachers were politically involvedand used the schools as platforms for their politicarbellefs. by indoctrinatingthe 'children fromP1 through senior secondary sdhool and by involving them in political activities in the later school years. Inthis regard, money and school materialswere said to be given or withheld by the Northin accordance with the"cooperation" of groups in the southern provinces. Money and books were available forteaching Arabic and subject matter in Arabic,but were said to be withheld forsubject matter inthe vernacular and English. As a result, parents and townspeople distrust the educational system, believing it to be a tool for theArabization of the South.

Because of the conflict, some parents would send their childrenonly to a "vernacular" school and others would not allow them to go to school atall. A large number of parents verbalized theirdismay at the increasingloss of language and culture by members of the youngergeneration.Many feel that this is happening because they believethe North is carefully orchestratinga plan to convertto Islas notonlythe.Southern Sudanesebut alsoblack Africans of other nations further south.

Thus, although on the surface thereseems to be a real desire for education among oldand young alike, there alsoseeas to exist a covert mistrust of education. Many of the Southern Sudanese claim thatthe Northerners are using the educational process as a mediumfor converting their childrento Islam. They also Claim that this process isdestroying traditional Sudaneseblack

35 PAGE 23

African culture, breaking down tribal andfamily relationships and distorting economic patterns upon which thesegroupsdepend. Although Hoppe's study did notdelve intothis matter in depthto determine the reality of what is happening regarding these claims, the factthat overhalf of thepeople expressed these feelings certainly indicates thata problem existsin this area.

A first step in dealing with this problea may be the recognitionthat political and philosophical differencesbetween the North andthe South have spilled over into the educational system. In order for a workable educational system to be developed for Southern Sudan, differencesmust be appreciated, appropriate prograas developed, economic needs shared. As Beshir (1974, p. 11) points out, new educational values for a new future must be hammeredout: "The problem! today Is what new values and attitudesare tobe injected into the minds of those enrolled in the schools, colleges and universitiesand how to do this...." (p. 1). The real hope of Southern Sudanis its people-- so much potential, flexibility and genuine concern.The bright spot is their deep desire for education -- education thatthey believe canbe used to overcome the difficulties of their present situationand the visions they have for their children and countryand the hopethat one day they will be able to direct the future of their land.

36 PAGE 24

3.0 THE LITERACY MATERIALS

Of the nine Role A languages, five -- Bari, Dinka, Lotuho, Moru and Ndogo -- already had orthographies developed for pea by aissionariesprior to the Civil Var (see Gabjanda & Bell,1979). The SIL personnel decided detailed xamination of these against a orthographies,which might have resultedin possible changes, for two practicalreasons. First, the IRL literacy officers assigned to work with themwere familiar with the existing felt they were adequate. orthographies and Second, SIL Was under considerablepresnure from the Southern Regional Ministry of Educationto show sone progressas soon as possible. Itwas, therefore, decided to forge ahead developing these languages which primers in were now designated "Cycle 1" languages.The project comaenced in Palotakain June 1977 andby 1979 trial editions four of the five Cycle 1 of priaers in languages (Bari, Dinka, Lotuho,and Ndogo) had been completed, pilot tested in .14 schools, and .revisedi Work on Meru had had to be postponed because the Sudanesesent to collaborate with the had great difficulty SIL consultants grasping the technology ofmaterials development. The completion date for trial editions of priners in kresh,Muer, and Zande, now designated "Cycle 2" languages, was scheduled forlate 1980 or early 1981 to lack of personnel, facilities due to housethe language officers, andthe need for further linguistic analysis.No work could be undertaken to on Shilluk due the unavailability of nativespeakers. .Theconstruction of additional living quarters.at Marldi enabled the project to be covedthere in 1979 and by 1980 the trial edition; of the Cycle 2 primers were almost finished. Some prellainary linguisticresearch on selected Role Blanguages the Jur Bell dialects, Jur (Mundu, Murle, Luo and Bongo) was carried outin 1977 and 1978. A six4ear timetable was worked'out for ihe production ea'trial editions of Cycle 1 and Cycle .2 primers, 'their field testing and 'revisionsand training of teachers.' the By. October 1979, primers hadbeen producedin Cycle 1 languages'and were being tried outin local schools. 2 primers began Preparation of the Cycle in Mareh 1980 at. Marldi. Their testing was tobegin in the 1981 school year.' Work on the preparation of theEnglish andArabic bridge materials, had just begun in.January, 1980, when the CALconsultants made their initial visit. 'At that' tie* schools were not in session,so the consultants focused on in extainition of the Cycle 1 materials and planning research strategy. future Thefollowing sections ofthis chaptercontain a description of the methodologyused inpreparing the Cycle 1 critique of these materials, materials, a recommendations for laprovement andother results of a seminar:held in Maridi in October 1980 during whichvarious changes to be incorporated in the Cycle 2 primerswere agreed upon.

3.1 METHOD

Themethod usedby theSIL group toassembledata andprocess it in preparation for writing the literacy materials is detailed in theManual of

37 PAGE 25

Literacy for Preliterate Peoples by thelate SarahC. Gudschinsky (1973). Gudschinsky, an excellentlinguisttrained in the tradition of Leonard Bloomfield, devoted considerable attention tothe subject of literacy. Her approach to theteadhing of reading does not drawheavily on any particular method presentlyin use, ratherit isderived primarily from her extensive personal experience in this field. The Manualis intended as a which can be applied to "cook book" produce literacy saterials anywhere in the world within a shorttime and it has guided SIL efforts for many years.Literacy Primers: The Gudschinsky Method in an updated and modified version of of the Gudschinsky system, authored by ErnestV. Lee (1982), an SIL linguist. The .preliminary stages undertakenin thepreparationof the described to us by the SIL primers personnelfollow the procedure set downin the Manual. As mentioned earlier Sudanese teachers assigned to the literacy project had previously taught thefourlanguages initially selected for concentration -- Bari, Ndogo, Lotuko and Dinka.-- using primerswhich had an orthography Weed on sylbols,developed at the Rejaf conference in which the 1926 at famous,,,Africanist, Dietrich Vestermann waspresent. The SIL Ungulate:worked withhthese teachers, learning as such as they could aboutthe languages and providing them with initialtraining in the Gudschinsky method. literacy The SIL, personnel had theSudaneseteachers write several hundred pages of.stories-in their languages.The only restriction placedon content ,of the the stories was that they 'be.typicalof the respective cultures. These stories then.constituteda corpus, which was thedata base for a frequendy count'Mede on all morphemes contained therein.The frequency count was arranged : In two categories -.7 content words and"functors" (grammatical morphemes). AWadditional frequencycount for all alphabetic letterswas made (Gudschinsky,, '1973, p. 55). basedon thecontentwords. Then a rough estimate of the test frequent sentence types was made. The SIL linguists then began tO coach Sudanese writers in the preparation of thematerials. The frequencycounts. forsed the basis fOrarriving at decisionson the introduction and sequencing of letters,word and functors.'

3.2 DESIGN OF THE MATERIALS

The pre-primers, each 110 pages in length, contained exercisesto develop the,students' ability to hold a penciland make some of the strokeswhich are components of the alphabet letters. A few sight wordsnecessary for the introductory -lessons in the primerwere alsotaught throughpictures. Exasples are shown in,Appendix 31.

He were ince:seed that the purpose or the primers was toexpose the students to all of the letters of the alphabet "in allof thecontexts they occur and to teach all .of :the basic functors." Itis a basiccontention ofthe Gudschinksy method that satisfying these two conditionseffectively equips a child toread withoyt hesitation the written formof any speechact, i.e., resUlts in the attainment of functional literacy. In fact astricter, operational definition of functional literacy for this project wouldhave been that upon completing the primer the student would be abletoread all of the stories in the post-primers. The level of proficiency required for this might begreater or less than thatattained by Sudanese children who have

38 , PAGE 26

completed a similar numberof yearsin schools were Arabic or English taught. is

One or two primers for each language were developeddependingon how much material needed to be covered priorto the post-primer. The format was always the same. Three types of lessons followsequentiallr: (a) Letter Drills; (b) Functor Drills; (c) Review Lessons.Each lesson contained some connected reading using.words introduced earlierin the lesson.

The-Letter Drill consisted of fiveparts. it began by displayinga "key" word, 1.0"-a-content word (a noun orverb) matChed with a picture. Then, in boxes, there were five component stages: (a) analysis, whichisolated the letter that was to be practiced; (b) .synthesis, whichwas usedto focus on the smallerbits within the unit, e.g., avowel.withinaconsonant-vowel unit; (c) an Identifidation drillt.which was supposed to provide the student withan opportunity' to pee :the "sameness"r of the new-letter4ndIdentify it; (d) a contrast drillahlik ha&thepurpose. of Contrasting thenew letter previoutly*udied with lettershich,occur in the same position.In'the syllable, and leYa'Word4uildintdrill, which built wordt with thesame letter studied in the wordland.previOusly studied letters. The-purpose here was to givethe student the.opportunity to see:thenew letterin larger chunks'ind "to tie-it to meaningful units."' An.example of a Letter,Drill'is shown inAppendix 38.

The,FunctovDrill; were Compoted of ttiree. parts (again, separated in boxes): .110 an analysis drill, which itolated the morphemefrom the broader frame prOVided*.the..sentenCe (the-stated.purpose, defined in tagmemiC terms, was to prOyide'an eite101efortocusing.On'thepOsitiOn of the aorpheme in its "slot"); .A0),anAdentification drill, which waatupposed to provide the opportunityfOr..00usiingon thevisual shape of ,the functor;and (c) a contriet'ArillOihick.,dootrasted thefunction in: a. sentenceframe with a previoUsiyAiarned fUnctoiin a7.sent:enCe.WhiCh,-as near as possible, was otheriitiqdentical Oncaagain,. aswait Vhe 'case' with the letterdrill, a short stret0h**nnedted-disCoUrse, usingentences with famillarwordsand the new4lunCtOr-vlaspresented on.the-oppOsitcpage fOr reading practice. An example:ora Functor Drill'is shown inAppendix 3C.

TheRevier,initsconsisted'of... three parts: -(a) a box in which syllable shapes, WhiOhWere-not always completemorphemes, were contrasted; (b) a list of words rpreirioutly. taught;- And (c) 'a boxin which- functorswere contrasted in sehtence.fraaes. Againt'a reading passige appearedon. the opposite page. Appendix. 3D'Ohows an example ofaReview Unit.

All teadher's editions of theliteracy materials cameprovided with a simple but effective teacher's guide written in the vernacular that outlined the teaching procedure for eachsection along with the appropriatepage of the pupils' book.

3.3PROBLEMS WITH THE CYCLE 1 MATERIALS

While in Juba and Maridi during theirJanuary 1980visit, the consultants had the opportunity to inspect all of the Cycle1 primers with two Sudanese

39 PAGE 27

linguists, Dr. James Dahab Gabjanda and Dr.Ushari Mahmud of the Instituteof African and Asian Studies at theUniversity of Khartoum. The SIL linguistic consultantswho had supervisedthe preparationofthe Cycle 1 primers, carefullyled the consultants through a page-by-pageexamination ofthese materials. TWo major types of problemareas were discovered.

3.3.1 PROBLEMS WITH ORTHOGRAPHY CONSTRUCTION

The decision to use the original orthographiesdesigned for theCycle 1 languages led to a number of difficulties.sincethese orthographiesoften violate an accepted principle of alphabetconstruction-- that the need to maintain a consistent shape for morphemes should takeprecedence over the desire to achieve a. I-perfect fit between sounds and symbols. This canbe simply illustratedin4merican Engliah,where.a single printed form of the past'tense morphemey is'used in.spite of the fact that there are three phonemic shapes, /d/ and" d/.' TheOrthographies devised for the Cycle 1 materials .. often violated' this principle bysymbolizing toomuch phonetic detail andrepresenting morphophonemic alternations that are not subject to ambiguity. The folloOing examples illustrate thesepoints. In Dinka the conjunction "and" isnormally pronounced /ku/ and hencewas transcribed as ku in sentences like (1):

(1) Akec ku ,Kon aci wel cop (name) and (name)pers./past 'guinea-fowlchase "Akec and Kon chased the/a guineafowl"

However, Wis,frequently written kuain sentences like (2):

(2) Eon aket wel kua aweec (name) 3rd pers/carrying guinea-fOwlandbird "Eon is carrying (the/a) guinea fowland (the/a) bird"

The a added to the'ku in example (2) represents theDinka consultant's correct perception of the initial vowel in the word."bird."What has occurred in this case is compensatory lengthening, a phenomenon quite coamonin many languages. Coming off the /4/91'111th is realizedlargely as lip rounding, the speaker begins, the articulation of the following, the /a/ of aweec,with the resultthat the latter it .lengthened. Thia would haveoccurred withany vowel, it just so happened that all of the words in pages 85through 91 of the Dinka primer (when itoccurs some 23 times) that followed ku began with What the Dinka an a. consultant insIsted upon carryingoverinto their orthography in all:of these cases wad theextra lengthening of vowel in the following word, hence theextra a is addedto' theku inonly' these instances. Predictably, .. the spelling always,remains mhen the following segment is a constant as in (1). The preservation of the formku as"and", which is motivated by the principle that orthography should facilitate the extracting of meaning, isvery relevant 'here. It makes no more senseto have two spellings forthis morpheme in Dinkathan to have two alternative like going and Ea in English. spellings In fact, errors of this kind are a violation

40 111 I Ii 8 3:: 1 ' % 'Pips 1 iiis 1 ia;1 ' ' /I ii 1"I 11 hi sPli iiti lit si a 11111 iiii ,,°'spi . I aalIasiail 1 I 1 a , 1 i il i wa. ii I li is 0 , li i' II iii ia iii PAGE 29

'and then Omoro is smokinghis pipe"

d. 1. Iwo Mahe oromo liana honyi ..> and(name) le hoeing in field his "Auks is hoeing in his field"

It &ohs 010 romo Luna honyi (memo) cesils soingbp hoe in fieldhis %mho going to hoe in his field"

. eriqo asoloq --) eriqo moloq meat baboon meat (of) the baboon

f. eriqo eyaya -- riqo hiyaya meat porcupine porcupine meat

One of the classic studies of howwords are stored In memory, Brown and Weill (110) produed evidence thatthe beginnings and ends ofwords.are Oritieel to their recognition and retention. As theexamples aboveshow, LOAD orthographyresults in frequent alterations of thebeginings of words, and bleed, 011 Many u three or tour such alternativesmay occurin a given autemee. Ilse thou!: these shape changesmay cause no serious problems tor *Mod readers, It 0000eivable that beginning readers mightbe confUsed by the trusted femme The problemve are hypothesising here isanalogous to what might MOW if ohildren learning to read Englieh first encounteredwords in their fhll fern, 41416, 01E, morr-anu, etc. and then labor hardto thee with the initial consonantor vowel removed, i.e., ther, seder, , Ste. Or course the only wayto determineif thisis a sirsiBUS iiiiiii:tIn the development of reading skillsis through controlled experimentation.Out such speculatiohland testing wouldhave been obviated by 1:11:11 the morphemes a constant shape and trusting in the native speaker to the phonological rules of hislanguage, i.e., make thedeletions and vomel changes himself.

The emd result ofmixing different levels intranscription isthat the orthographies devised terthe Cycle 1 languages vacillatebetween the overly phonetic symbolisationtypical of transitional, learning alphabetslike the 1.t.a. (international teaching alphabet) and'aphonemic representation that is mot optimally efficientfOr adult reading. This compromise orthography does mot optimally serve tither of thetwo audiences-- literateadults or beginning readers-- which might wish to make use of it.

A reiated issue Uhich Should probablyhave been looked at carefullywas *ether or mot some homophonous semanticconcepts might not havebeen more readily identifiablethrough the selective use of diacritics. Tone is not represented in *logo spelling; thus, those morphemes with different meanings. distinguished solely by this feature must beprocessed with careif reading mansions are to be avoided.TO cite one example: theletter a represents the oopula,e.g., I a no, "Kip (antelope)is an animal." An alternate pattern Imig is umed wi adjectives: A niki a no, "the animal is good." Out sembs initially,a serves as thepassivemarker for immediately following verb,e.g., 4 na no nil"the animal was killed." The different spoken realisations of the single lettera are apparently distinguished by the tone, but thisLe not carried overinto orthography. Since the primer

,. 42 PAGE 30

contains alarge number of morphemes like this, itis easy for areader to assign an incorrect meaning to a partially processedsentence. For example, note thatin (5) and (6), the first two morphemes are print. indistinguishable in

(5) A ku na Passivepresent/progressive shoot marker aspect marker 40 "(It). Is being thot"

(6) A ku no (It) ishide/skin animal "(It) is the hide of in animal"

There it a gooddeal of basic researchsupporting Goodman's (1967) "psycholinguistic guessing game"model of reading, whereby the what is reader predicts likelyjb occur', next.'in thesentence based on thematerial already assimilated..Comprehension odours ifthe-Predictions are fulfilled butbreaks down.whenever 'unintiebeed:materialOrops- up.Now it is very clear that, given :-theibsence ::Of;t0ne. 'markings in-- the Ndogo orthography, abeginning reader-sight-Well.Startproceisingeither ofthe two sentencesshown above with'the'_unintendedieaning." IfileAoes.not notice .the difference betweenna and:na:Womitassign:a:completely"....incorrectreading to the sentences. If he dOesilotiCethe,..differenda:he will,have.to stop, backtrack, andfigure out the73O6e6d-OfconfUsion.Beginning readers are notoriouslypoor at resolving theiekindsOf ambiguities, frecitiehtly becomingtOtally confusedand disoriented,when.unintehtionallymisled in.this manner.

The possibilitir of confusions of this type occurring in Ndogoshould not by minimized...We tound a number ofhomonymicmorphemes in our inspection of the first primer, e.g.', mi = (a)' "x is/exists", (b)habitual 'progressive aspect marker,(o) "skin, hide"; se = (a) "outside," (b) "fish"; gi= (a) morpheme indicating completed action,-(b) intensifier "very", (c) verb "come";ta = (a) (b) "and"; ka .= (a) "just", (b) relativeclause introducer "whiCh/Who",.(c) "is/ariw; ti =JO"with% (b) "for the sake of" benefactive with tacoL (c) "in", (d)."at" (as in "look at" something).Of course not all of-these-give rise to asbiguitiest'sincesone meanings are restricted to the occurrence'of that,morpheme in'a given position in a sentence. But there is sufficient overlap for :thepoSsibilityfor -erroneous readings' ofa given morpheme: to justify, considering the use of selective tone marking. For exempla-the use of a. diacritic for the passive 'marker, e.g., J. would have effectively naturalized the potentialconfusion alluded to in examples (5)and (6). .-

3.3.2 PROBLEMS RELATED TO VOCABULARY CONTROL

A far more serious problem than the infelicities discovered in the Cycle 1 orthographies had 'to do with the introduction and repetitionof new words. The SIL/IRL writing teens had adopted a policy ofdrastically curtailing the number of new words that could ippear in each consecutive lesson. Furthermore each new word was recycled a minimum of.five times in five successivelessons

43 PAGE 31

after its initial appearance.We were told that the rationaleunderlying this policy was that recycling this often promoted word recognition andretention. This wordrecycling policy had twoundersirable effects, the first of was the which occasional creation of unnaturalpassages, whereadjacent sentences had no logical connection. An excellent example of this is found on page 71 of the Ndogo primer which is repeated in its entirety in (7)below. (7) Nda ce ku iri ka co to. plural snail pres.-prog. go/crawlaround down.

Mu niki tindula, Mu niki a 1 tree good for they nottree good are something

ti nda mu, taco ndu ku for plural bird because theypres.-prog. fly

ka co tari. Nda ce o la, ndu round sky, plural snail fly not, they

ku iri ka to. Nda da mu la. pres.-prog. go/crawl over down. plural chopwood not.

"Snalls.crawl around on the ground.Trees are not good for them. Trees are good for birds, because they fly up in thesky. Snails don't fly, they go around on the ground. They don't chop wood."

The final sentence, which is clearlyincongruous in light of the preceding context, was attached because the word "chop" needed to berepeated one more time to conform with the constraint onthe repetition of new words.There is a growing body of research, e.g., Glenn (1978),Kintsch andVan Dijk (1975), Handler and Johnson (1977), Stein (1979), Stein and Glenn (1979), which suggests that the word recycling policymay be counterproductive if it results in testa like the above, which distortepisodic structure, since comprehension and retention of contentappear to be highly dependenton the stories. sturcture of

The constant repetition of a limited numberof words also produceda large number of.totally predictable sentences. Although there is no evidence thata high degree of vocabulary repetition is' optimal for reading,one study by Carterette and Jones (1963) suggests that it may be extremelydetrimental. Seeking to determine how much vocabularyrepetition ispresent inreading material that Children prefer, these two researchers asked librarians fromall over the United States to.providethen with a list of thosestories most popular with children. This popularity ratingwas determined according to two criteria: those story books which were taken out most frequently andthose stories which children liked to have read instory hours. Three levels of stories were considered in this study: (a) storieswritten for kindergarten up to the second or third grade; (b) those written for Grades2 and 3 up to Grade 5, and (c) those written forchildren from Grades 4 to 6.

The study turned up a number ofinteresting facts.First, there was little evidence that the lengthof words wasuniform at any of the three levels. Long words and short words seemed to be includedregardless ofthe level. PAGE 32

However, there was a distinct increase insentence length as gradelevel increased.Nest importantly, however, whenCarterette and Jones compared the preferred stories at the first level (kindergarten togrades 2 and3) with comaercial readers dezigned for thesegrades, they found thevocabulary repetition rate in the preferred stories to be considerably lower. This would indicate thatchildren who are beginning to read may very well bebored by commercial readers because the rateat which words are recycled inthen is so high.

Soneof the Cycle 1 aaterials provide a graphicillustration ofthis criticism. One Sudanese graduate student whoassisted Cziko in preparation of the the Bari literacytestswas sufficientlyimpressed by the repetition rate of words and sentencesin the Bari primerto remark that the book appeared extremelrboring .to himand might well be perceivedas such by the children .for whomit was designed.However, the constantrepetition of the same vocabulary poses a far greater danger to the development ofeffective reading skills than was apparently recognized by the SIL staff. In order to determine with reasonable precision just how redundant a Cycle 1priaer could be, Cowan selected one.-- the BariBook 2 -- and did a vocabulary,count of the words used,in sentences in the first 35 pages (theprimeris 75 pages long). Teble.3.1 shows the total frequency of the.words used in the first 35pages'.

The high frequency counts for thetwonames can be ignored, since Jade and Poni are the aain characters in,the stories and hence appear in practically every sentence, But:: note that only 10nouns and 5 verbs are used in the 35 pages.- An.inspection Of.the.nouniandverb talliesreveals the extraordinary repetition of vocabulary.' TwO'nouns, tapini (guinea fowl)-and kavate (sweet potato).appearAlmost.as Zanytimesas, the name Jada. Also, two verbs, aria (cook) and .myesu (eat) have extremelyhighcounts, 26 and27 respectively. (The low-'Countm for lala (wash) and di (say) do- not reflect thefact that these two'verbs are not introduced into stories until page29.) Given these two factsit should be obvious that praCtically every page of the firsthalf of this biz* contains one of eightsentences: a. Poni aderja tapini "Poni is cooking theguinea fowl" b. Poni aderja kayata "Poni iscooking the sweet potatoes" c. Poni anyesu tapini "Pont is eatingthe guinea fowl" d. Jada anyesu taploi "Jada is eatingthe guinea fowl" e. Poni anyesu kayata "Poni is eatingthe sweet potatoes" f. Jada anyesu kayata "Jade is eatingthe sweet potatoes" g. Poni koJada anyesu "Poni and Jada are eating tapini guinea fowl" h. Poni koJada anyesu "Poni and Jade are eating kayata sweet potatoes"

There is considerable research which shuls thattwo factors assist children in decodingwords. Thefirst is the overell frequency ofthe word in the

45 PAGE 32a

Table 3.1

Vocabulary Frequency of Bari Book 2, Pages2-35

Word Frequency Nouns

Jada (name) 39 Poni (name) 56 tapini 'guinea fowl' 33 kayata 'sweet potato' i2 lokore 'meat, food' 9 muri 'dyker, antelope' 4 kima 'millet, grain' 14 koropo 'leaves, beans' 11 tuki 'jug, pot' 9 kolu 'cola nut' 6 rabolo 'banana' 6

Verbs tatu ¶:hoot' 11 derja 'cook' 26 nyesu 'eat' 27 lala 'wash' 2 di 'say' 6 kata ko 'have' 9

Prepositions ti 'of' 13 kata (i) 'in/inside, In the place of' 12

Conjunctions ko 'and' 26

Adverbs dika 'in the morning'

46 PAGE 33

language. Children have less trouble withvery familiar words. The second factor hasto do with thebeginning andends of words.Thesecarry more information than the middle ofa word and hence children mayaake a guess at what a word means based on partial recognition-- i.e., the first few letters, the last, or both the beginning and the ends.

Now, all of the words used in th6 Baripriaer arevery familiar. It undoubtedly doesn't take very long before thechildren realize that one of the 8 versions (a -h) is going tooccur on every page. Thisbeing the case, they can adopt a strategy,which will allow then toguess any sentence with 100% accuracy (and tnus perfora wellin class)by attending to very little information in the print. For (a) through (f), the children haveonly to look atthe first letter of the first and last word todetermine the subject (either Poni Jade ) andthe object(either tapingi kayata). Similarly, by looking at the first two letters ofthe verb, they candiscover whether the subject is cooking or eating.

One cannot escape the conclusion that thepolicy of recycling new vocabulary five tines in five successivelessons carries with it thevery real danger that the children are not learning the word-attack skillsthe letter drills are supposed to teach. Rather, this policy is more likely topromote an unproductive analphabetic strategy-- a kind of trick which will even fool the teacher: "Look at the firsttwo letters of each word andyoucan tell what the word is." In fact, there was some anecdotalevidence that children who had been exposedto the Cycle 1 materials were notlearning to read, but were slimly memorizing thecontents ofthe very redundant stories. Wanda Pace, the SIL consultant who helped developthe Ndogo andDinka materials, stated thatiin spot tests with Ndogo sentences,the children would seize upon one word, Ahe name of one of the charactersin the primer andinstead of reading through the actual sentence, would simplysake up sentences which were actionsthat character typically performedin storiesthey had memorized. Obviously thereis no wayof determining whether the children are actually learningtoreadunlessthey areconfronted periodicallywithpassages containing words they have never seen before.

3.3.3 PROOFREADING

The proofreading of the primerpassages left much to be desired.We found errors whiehdestroyed the logical developmentof a prose passage, like the one on page 83 of Lotuho Book I. In this story, Achaha, one of the main characters in thisprimer, discovers some porcupine meat thatbelongs to him at Ihuma's house.Enraged, he draws his knife and Ihuma, obviouslyfrightened by this gesture, runs away to Omoro'sfield. There Ihuma finds Omoro sitting under his thatdhed sun shelter eatingpumpkin.Thefinal sentence of the story, which caps thisentire sequence of events is: attati inyi ifa jyo Omoro. "And then- they (both) runaway from Osoro." This obviouslyflakes no sense in light ofthe preceding events and onecan easily imagine the consternation of the native speaker when theconclusion is read.

Sometimes the errors were due to the fact that the SILlinguists hRd not had the opportunity to gain more than a casual coamand ofthe language with

4 7 PAGE 34

which they were working. For exasple, the exercises onpages 62 and 70 of the Bari priser are designed t43 teachthe students when theorthography signals vowel harsony, but some of the words selectedto exemplify the application and absence of this rule are not correctlyallocated.

3.4 THE ENGLISH BRIDGE MATERIALS

In those rural schools where bookshave been available for teaching English in P1, the most widely used text was FirstYear English for Africa, Part 1 (Speaking) and Part 2 (Reading), by F. G. French. The SIL consultantin charge of English materialsfelt that this book was unsuitable for the Sudan since it contains references to machine made products andpictures of things unknown to Sudanese children.

Mbst rural schools have few facilitiesbeyond a blackboard. Many of them are nothing more than open air classroomswith crude benches for the children to siton., For this reasonit was decidedthat theLevel 1 (first year) English materials should be simple anddurable enough to beused for several years. The Level 1 materials were to consist of 17chartswith pictures'of subjects around which oral lessons willbe built.The topics selected for the charts were; (i) greetings, (b) animals, (c) body parts, (d) people; (e) family sesbers, (f) "action" verbs, e.g., sitdown, stand up, squat down, cook, eat, drink, (g) roods, (h) more action verbs: walk, run, talk, look 177(1I--terms that pertain tothe school, (j) household items, ---(E) the numbers 1 through 6, (1) expressions pertainingto life at home in the house, (m) life at home outside the house, and (n) the numbers7 through 12. A single chart was to be used for Lessons 15 and 16, the topic being thingsone sees on the.road. Lesson 17 teaches the upper and lower CaSe letters of the English alphabet.

The goals of the firstyear course were modest-- the childshould know about 100 English words and the alphabet. In Naridi the consultantswere able to examine the teacher's guide for theLevel 1 English materials. It contains two pages of instructions to the teacher concerning general procedures(e.g., always speak English in class), basic considerations foran oralEnglish course, (e.g., provide a good model of spokenEnglish for the students) and helpful hints (e.g., use songs andgames to make the instruction interesting). This is followed by a demonstration of themethod tobe used in teaching English from thecharts, which is presented in the form of a sample lesson. Finally, there are lesson plans for all 17 charts and a chart showing the strokes to be used in making theupper and lower case letters of the alphabet. The guide seems to cover all of the importantthings a teacher should beaware, of and it is written in clear, explicitlanguage.Whether it needs to be made more explicit or simpler to be of realbenefit issomething thatwe were unableto ascertain since therewas no opportunity tospeak withEnglish teachers. We were told that the work on the chartsWILS in progress. It Was being impeded somewhatby the fact that theSIL expert on ESL was in Maridi while the artist who was illustratingthecharts lad at the IRL office in Juba.

48

.. PAGE 35

The SIL consultant was also workingon an elementary reader patterned after the popular text used in Kenya, Hallo Children! (The Progressive PeakEnglish Course, Book1, Oxford Univerial-lEWZ77- Aplanfor Level 2English materials has been completed. The projectedlessons willcover possessive pronouns, demonstrative pronouns, plural forms, negation, time days of the week, conjunctioas, concepts, the colors, sore action verbs, the verbhave, the adjectives had, good, go, down, hot, cold, the verb want followedby a noun phase and for/to complement, assorted .prepositions, indirect objects withthe verb an, teacher expressions suchas listen to and talk.to, the verbsbrim, come, take, the past tense with the verbs grja, cook, 102, close, walk, and kick, and the irregular past with the following verbs that undergosuppletion: eat, sit, go, come, king, andsee. Sentence building activities using on flash cards will be words introduced after completion ofthe aforementioned material.There is a third year reader planned, which is intended toprepare the 'children for the Primary 4 materialsthat the Ministry of Education is using.

The SIL person in charge ofthe Englishprogram seemed to be competent, dedicated and enthuaiastic. The idea of basing the materials on situations that the child is familiar with was a good one and the projected schedulefor completing various phases ofthe materials developmentappeared realistic. to pe

3.5RECOMMENDATIONS FOR IMPROVEMENTS

In addition to the problems detailed inSections 3.3.1and3.3.2, examination of the Cycle 1 literacy primers prompted the consultants tomake a nusber of broader recommendations for theirmodification or revision. It was noted that the format of the Cycle 1 materials gave the superficialimpression that the childis learningto readbecauseheis graduallyexposed sound-symbol to correspondences and grammatical morphemes. As the child completeaeich lesson and moves on, a sense of progress is felt. Due to the high repetition of the vocabulary, one mighteven hypothesize that the child could learn every word in theprimer by sight alone. What is likely to happen is that the child is actuallybeing trained to recognizemany of the morphemes he will encounter in the post-primer. Of course everything thathe retains from the primer will assist him in the decoding act, but it cannotbe strictly claimed that'this approach is imparting skills which will enablethe child to decode new texts composed of elementshe has never seen.

Afundamental assumption of readingtheory which is supported by considerable empirical researchis that the readerextracts information from texts using several sources of availableinformation -- knowledge ofmeaning, grammar and correspondences between writing andspeaking. Thereader undoubtedly brings these sources tobearon the decoding task simultaneously, sometimes emphasizing one or the other skilldepending on the natureof the task. Most innovative reading methods provideopportunitiesfor showing the child that he say exploit all of these sources. The Cycle 1 materials, in contrast, emphasize only one of thesesources, sound-symbol correspondences. It ishard to believe that theFunctor Drills doanything morethan teach

49 PAGE 36

sight recoinition of the grammaticalmorphemes; they certainly do not teach the child to use his knowledge of thegrammar to assist in decoding.

The Cycle 2 materials are deficient in thatthey provide no opportunity for the child to realize that meaning can be extractedfromclues which are present at various levels: (a) partially available visual informationwithin the word asin g_ss ng w rds w th t v_y_ls; (b) grammatical knowledge (e.g., using knowledge ofgrammar to recognize that a new word is anoun, verb *or adjective, etc. and then making an intelligentguess as to what it may mean based on the rest of the sentence or the preceding sentences, and; (c) meaning (e.g., given the broader contextthat I am looking at the Nile River, if the next sentence is: "Suddenly I saw a f move below the surface" you realize that the possibilities of whatthat could be are extremely limited). If the primers areto actually train children toread, they must incorporate opportunities for the Children to becomeawareof these kinds of clues. They should also contain passer's whicharein advance of the children's abilities in order to provide practice in usingallthree levels of information in decoding.

3.6THE READING WORKSHOP

The above recommendations for improvements in theliterary materials were viewed favorably by John and David Bendor-Sanuel (two SIL administrators),and at a meeting held in April, 1980, attended by G. Richard TUcker, Director of CAL, thetwo CAL consultants .and theBendors -Samuel, it was suggested that COwan provide a workshop on reading theory and research for SIL/IRLpersonnel working on the Cycle 2 materials in Maridi. John Bendor-Saauel proposed that the workshop take place soietime afterthe report of theconsultants' first visit (COwan and Cziko, 1980), had been circulated inSudan, but soon enough so that any innovations which mightcomefrom theworkshop couldbe incorporatedinto the Cycle 2materials. It was emphasized that SIL/IRL personnel should not feel obligated to incorporate any suchinnovations, if this would pose a serious setback tothe timetable for the Cycle 2 materials. Two tentative dates were proposed forthefirst workshop-- August,or late fall, the most acceptable time tobe determined throughcorrespondence with the SIL Director of Literacy, Julie Van Dyken and JamesDahab Gabjanda, who was to continue as CAL's in-countryresource person.

The workahop eventually took placeat the. IRL Training Center in Maridiin late October of 1980. Due to illness, Dahab Gabjanda was unable toaccompany COwan to the South. While in Juba, COwanwas able to confer with Eluzai Mop Yokwe, a faculty member at the Universityof Juba whowas assisting Cziko in the preparation of reading testsfor the Bari children.

Cowan arrived in Naridi on October 20th, 1980 and conferred with JulieVan Dyken. It was decided that the workshopwould coamence the next day and that only SIL personnel and their IRL counterpartswould take part. There.were thus twelve participants: eight SIL consultants- JulieVan Dyken, Jocelyn Clevenger, Wanda Pace, Claudia Scott, Rick Brown, Marian Farquhar, Alice Van Bergen and Rick Sampson --and four IRL counterparts: EdwardB. Mandeson, Vice Principal, IRL; John Baptist Asan, Patrick Rasadan Ladu and Daniel Ladu.

50 PAGE 37

Regrettably, Job Dharual Malou, theDirector ofIRL, was not at Maridi during the swifter.

The workshop began witha series of lectures which treated reading theory and aspects of reading research. These were intended to providea basis for discussing bow the literacy materials could be improved. Cowan's intention vas to move from a lecture format toa general discussion of the issues and then, raised if tise permitted to work-sessionswherethe participants would attemptto try out suggested innovations using Cycle 2 materials.As turnedout, it this INIS more orless what happened. WhatfollowsiS an abbreviated summary of the lecture.

3.6.1CURRENT READING THEORY AND RESEARCH

At present there are three major theoreticalapproaches the process of reading. to understanding The first of these,usually referred to asthe "bottom-up" approach, bolds that bothin learning to read a firstlanguage, as well asin mature reading, there is a Unidirectional flow ofinformation, beginning with visual, i.e., graphicinput and proceeding through.higher-level stages, 048., translating symbols to sound-units' (at the level of thephoneme or eyllabic unite), which are.in turn . amalgamated into larger units ofsounds (words or morphemes) which ere. recognized as meaningfulconcepts. Scholars associated with this-theoretical view of how reading comprehensiontakes placs are, Bloosfielt(1942), Bloomfieldlt Reinhart (1963), Fries; (1963), (1965)i. and .4ough (1972). Gibson Methods for 'teaching readingwhichtake the "bottom.4", Sieves basic, havebeen proposed by Blooafield (1963) and Fries t al.(1960. Their Y'predosinantfeature is a heavy reliance seek .to inculcateHappropriate on drills which symbol-soundcorrespondences forgraphemes, often at the expense Of Meaning. A good example of a basalreader series which eabodies the Bloomfield-Friesapproachis the Sullivan Associates "Read and Think" Series (1973),.publishedby McGraw-Hill.

The letter-by-letter processing andsubsequent amalgaaation into meaningful unitsimplied bythe "bottos-up"view of reaJing has beenshown tobe impossible. Tachistoscopic experiments in perceOtionhave demonstrated that although the eye'can perceive sorethan five units, it cannotregistera larger numberthan this. In other words, thereis a severe liaitto the asount of saterial which tan be inputfrom the sensory store into emory.. short-term .Obviously reading cannot proceed,given thisrestriction, ifthe units to be registered are graphemes or' even syllabic units composed oftwo grapheses. In language which containslong, polysyllabicwords, like Lotuhoolhe neophitehreader willpave forgotten what he has .readby the time he reaches the endof the word he is trying to "sound out." Thisexplains why children who have been taught to "sound out" using methods like phonicscannot handle longer words,and why theysisply lose track of the meaningof an entire sentence when they encounter them.

The only solution to this problem, one which permits reading tooccur as we know it, is for the reacier to input higher levelunits-- chunks composed of many lower-level units. Miller's (1956) classic article graphically demonstrates that thistype of recoding isessentialfor alltypes of

51 PAGE 38

processing, especially reading, (and recent research, by)4cConkle, 1982, has clearly establithed that the wordis theminimal unit perceivedby skilled readers).The obvious pedagogical isplication istbat any method whichkeeps the learner atthe lowest level of processingand does not provide him opportunities for bringing higher with levels of information to bearon the task, runs the risk of teaching him just to"bark out the Sounds," but to read. not to learn

The model of reading which envisions the use of largerunits of information has been referredto as a "top-down" theory. According to thistheory, the reader rarely makes use ofan intermediate speech code; rather,he is sensitiveto, different kinds of "clues" which facilitatechunking. As portions of the.printedpage are takenin, the reader posits the as yet unprocessed text. hypotheses about Chunking occurs because thereader uses natural properties of language suchas redundancy and syntactic structure. enables the reader to identify Redundancy largerstretches of material,even need not register all of the though he individual units of whichit is composed, e.g., all of the letters in word. Hypothesistesting isfacilitated bythe speaker's knowledge ofgrammar, which allows him to reduce possible kindsof information the numberof which can occur sequentially(Shannon and Heaver,: 1949). For example, the probabilityof. a determinerlike thea occurring immediately after the in- a text is virtuallyzero. Indeed, the possibility of some words occurringtothe exclusion of others stretches of '.11scourse. increases over Ihe top-down theory ofreading is most often associated withXfoodman (1967). Itis sisilarin many respects to Stevens-Halle (1967) analysis-by- the synthesis model of speechperception. The constructive proóess.proposed byRyan and Semmel (1967) andSmith's (1971) cue sampling process may be classifiedat top-down theories of reading. There it, empirical' evidence that beginning readersexhibitbehavior predicted by the top-down theoryof rea:ing. Veber's of children s (1970) oft-cited study oral reading.errorsindicates that, at the earliest stage of learningto read, children are sensitiveto the structurAl sentence they are attempting clues of the tolecode. She foundthat chlidren'serrors which resultedin ungrammatical sentences displayed great similarityto the graphic.configuration of the word they did not know.Children whodid not correct.errors which resulted in ungrammatical sentenceswere poorerreaders. consistently Similar results have beenobtainedby Neville and Pugh (1976), with diff,erent:types of close tests.They found do not avail ticiaelves'of that poorer readers contextual. confirmationand frequently "give up" whenever they; eticounter a wOrd that they do not know. Although comparative studies arerare and far'from conclusive, at least one indicatesthat teaching nethodology can aftect readingperformance. Elder (1966) found that children'who had ,received Scottish more amilytic word-attack trainingthan a group of American children, whosetraining had emphasized tryingto obtain meaning well as decoding, made as more errors that changed the meaningof sentences. The pedagogical implication of studies like these areclear. The methodology usedto teach Children to read should emphasize thatreading is meaningful and should 'not focus solely on developing theability to sound out unfasiliar words. The danger of relying tooheavily on teaching the child analyze wordi say, as Gibson to and Levine (1975, P. 238)have putit, cause beginning readers to becore "stuck"at thephase of handling only the information..." graphic Good reading materials will contain exercises which call the

52 PAGE 39

learner's attentionto theuse of differentlevelsofinformation-- redundancy within the orthographicsystem, syntactic and semantic information -- in such a way as to help him reduce uncertaintyat variouspoints in the text.

Recently reseachers have turnedtheir attention tohow the structure of discourse say affect reading comprehension. Scholars such as Rumeihart (1975, 1977) have proposed (the.notion of a "story grammar" i.e. specificprinciples by) which any type Ae pure passage- descriptions, explanations, jokes, paragraphs --Are .organized. Rummelhart hasarguedpersuasively thatthe organization a "grammar ofa text is extremely important for comprehending and recalling the information it contained in it. If this is true, it is quite important for the.design Of good reading materials, since it would implythat stories which deviate fro' the story grammar'Are harder to understand.The researckof Maras (1978): andThorndyke (1977)demonstrates that adults do have trouble recalling stories which violate a basiccanonical form. Kieras present 4dulte.with: paragraphs,all of which werecomposed of sentences containing a.single -preposition. These werescraabled in fourdifferent orders And pretented to:subjects one sentence at a time by a computer. At the end of the'presentation, the4ubjects were required to write.down the storyas accurately as j.they could and-later to write summaries. Thesubjects always did better'Wthese 'tasks When theparagraph had a "classic" topic eentence organization.of f011Owed by a.series ofsentences'linked to one another in terns orthejnformatiOn previoinplysupplied.Thorndyke has performed similar experlientir iittUonget storiescomposed of morethanone paragraph and obtaineCilmilar-reitlts, proving that the internal plays a Aajor structure of. the story comprehension. Children's sensitivitytostory structUre andits.impOrtance fOr'comprehension and retentionregardless of cultural-background' his been demonttrated by a number of studies: Handler, Scribner Cole and:DefOrest (1979),Handler and Johnson 0977), Stein0979), Stein.an;Glenn (1979) and KezworskY, Steinand Trabasso (1982). The pedagogical implication of this research is that connectedprose in beginning reading material must, if the purposeis tofacilitatethe acquisition of reading,tkills,. be governed by conventions, of coherence and naturalness. Pedagogical practices that lead tothe production of illogically structured sequences of sentences likethose described above are to avoided. be

Rased on the major points raised in these lectures, Cowan thenrecommended the following additions and changesin the format of the literacy.materials:

1. The Cycle 2 literacy materialscontain no provisions forteaching children how touse graphemic, syntactic and semanticclues inreading. Exercises toteach these clues shouldbe developed. Cloze proceduresare widely used for this purpose. The workshop participantswere provided with a taxonomy of close procedures and examples drawnform commercialreaders. A few adaptations using Lotuhowere also discussed.

2. The policy of recycling new vocabulary, which frequentlyforces writers to distort whatwould otherwise be a natural-sounding, coherently structured . story and defeats the whole purpose of the letter drills, should bemodified. PAGE 40

3. All non-meaningful elements of theexisting format should be changedto emphasize that reading is formeaning.

4. In order to create more variety and interest, emphasize thatreading is pleasurable and meaningful and providethe necessary reinforcementwhich will sustain .the proficiency gainedfrom theprimers, a separate story book be prepared. should This should contain stories about childrenand the way they interact- in the life of their village. Thestories will be continuous, running from 6to 15 pages. Each 'page would containan illustration of the events described in the three lines of printon that page. Writers were to carry-out.- simple research-- observing children at -play andaskingthem to talk about: what they regularly do and enjoy doing-- to gainaaterial from whiCh-stories miy be'created. Vocabularycontrcil would be relaxed in these , stories, which were, in- part, intended as a feedback measurefor the teacher 'to check.the Children's progress. The story Wok might be.prepared in the futurtat ''.the'same time the primers are written. Itwas hoped books might stillbe written for that story the Cycle2 materials. Examples of story books like those envisioned in thisrecommendation, which wereprepared for Hausa, a language spoken in Northern Nigeria, are shownin Appendix 3E. 5. As aenticaed by Cowan and CzikO (1980, p.25), the overalljustification fOr functor drills- is,not clear. In an unpublished manuscript,SIL linguist Ernest Lee :implies:that thepurpose of thefunctor drills is chunking. to facilitate UnfOrtWiately he is unable tosupply any compelling argumentsin support.ofthis implication. It we; suggested that the functordrills be orgapiZed 'on the .:baals of . semantic criteria rather thanon theirpresent syntactic' (Tagmemic,-"slot -filler")basis.One could devise an organizational schema composed or sepantic.:categoriee,e.g., aspects, directional, positional morphemes, temporal expressions,etc.

6. Inliterate societies, children approach their initialinstruction in reading_with i good deal more pre-reading trainingthan is afforded by the SIL/IRL preprimerb. In view of the complete lack of Written materialsin rural homes, which mightprovide reinforcement and generate reading, it .was.redommended some interest in that the existing pre-primersbe expanded, the pre-reading instruction be extended and an alphabet book beprepared for each language. _Exapplevof-rebus materials usedfor early pre-reading trainingin English were presented and theadvantages of-this approachwere discussed. At this" point.the reactions ofthe participants tothese recOmmendations were solicited.- The general response wasfavorable, but it wasfelt that considerable disdussion'would .berequired to determine which could recommendations be implemented ind what form thiswould take. Inorder to shutting down the groups woring avoid on the Cycle 2 primers, it was decidedthat a smaller group...would discusswhat the recommendations entailed in terms of. format changes andthen try them outwith Cycle 1 and2 languages. The discussion group Consisted of Julie VanDyken, Joycelyn Clevenger, WandaPace and Claudia Scott.Edward Nandeson was invited to participate but hisduties were too pressingtopermit this. Cowan waspresent .only toprovide additional information and clarification of any pointa madeearlier. It would be impossible to describe the intense discussionswhich took place during the next two days. These were admirably led by Julie VanDyken and sometimes went on well into the night. This approach was to examinein detail

54 wee 4.411411,71"if

Ate V.41 PAGE 42

reinforce the idea thatthe symbols on the page havemeaning. letter drill A sample of a lesson, devised for lesson 13of theNuer primerby Joycelyn Clevenger la shows in kpendix3F.

3.6.2411 22111VALIII

In accordance withthefirst recommendation made above,a new set of xercises was added to provide much neededpractice in using the different levels of Information in a text that canfacilitate comprehension. This innovation, vas to appear as a separatepage of exercises afterthe letter drills and beforethe story, was unanimouslyendorsedby the workshop participants. In accordance with the resultsof Weber's (1970) research, it was decided to provide a gradation of difficulty, starting with syntax,which lathe cluethat children are most attentive to. For lessons 1-25, the context clues would focus on syntax. Examples of Nuercloseexercises designed for these units by Jocelyn Clevengerare shown below. Notice that the choice of the correct word for the blank is syntacticallydetermined.

NenDcco cc rect

Choicest kal thal yieer

HanDecn she fish

Choices: compound cooked river

Units 215-50 would contain items like the one.shown.Here thechild must choose the most semantically plausible itemto complete the sentence. Blanks would now appear in the middle of the sentence as well as at theend.

GuanDeeg cc guec Ice mac.

Choices: yicth may bir

GuanDccg giraffe with gun.

Choices: speared fished shot

Lesson 51andthose following wouldcontaincloze items wherethe alternatives are graphemically similar, but theappropriatechoice is lip:tactically determined. The childwould now be forced to payclose attentionto both the spelling of the wordas well as contextfeatures. primer. The following example was devisedby Joycelyn Clavinger for Lesson 54 of the Nuer primer.

JAc cc luaak. Choices: rig r5g rig

Jzic he to barn. Choices: seat enough run

Guidelines for preparing close items and the teachingprocedure forthis section which were developed at theseminar are found in Appendix 30.

56 PAGE 43

It was felt that all of the words taught in this section should,initially, be -wordsthe children have been exposed to. Fromlesson 50 on, the alternatives mightbe built words. This new section would notreplace the existing Word Attack section.

3.6.2.1.3SUPPLEMENTARY STORY BOOKS

Recommendation 4 was unanimously endorsed. It wasdecided thatthe preparation of separate story books for the Cycle 2 languageswas possible, but that thiscould not be undertakenuntil after the Christmas story books would.have break. The a greater emphasison children's activities and the languagelnthem would contain examplesof nonreported, i.e., and questions. quoted speech It was decided that the childrenshould be introduced to the story book upon completion of therevisionsection of lesson 25 and that they should read a new story with eachnew revision lesson.

There was considerable discussion regarding the vocabularycontrol in the story books. A sliding scale was proposed: in lessons ?5-5-, 5% of all of the words may be unfamiliar; in lessons 50-75, the percentagemay be raised to 7%; in lessons 75-100, 8% and from lesson 100 on, 10% sight be permitted. It was finally decided to use the first setof story books asa test to determine what a reasonable figure for eachblock of lessons would be. The current teachingprocedure for the primers wasdevised appropriately for the story. bookt. The teacher wouldtell the students toread the story silently and then questions wOuld be asked about the content ofeach page. If the students could not answer the questions, the teacher wouldask a student to read aloud as much.of the first sentence as.he can. (The teacher would first refer to the:picture to set the scene.)The teacher would encourage the student. to Use the syntactic andsemantic clues available todecipher any words heis-having problems with. After the entire story hasbeen worked through in this ;tanner, the teacher wouldread thestory aloud. Then the teacher would read...each sentence. Individualreading aloud would follow. It was suggested that the class might be split into twogroups, one reading the primer and the other, the storybook.

One ofthe Most valuable suggestions made at the workshopsconcerned the use of-the story books to achieve a doubleimpact on the literary problem. Several participants noted that students willoften keep the primersat the end of:. the school year, even atthe riskof a fine. In areaswhere this occurs,- adults have been Seen attemptingto learn to read fromthe primers. It was suggested that if the Ministry could be encouraged to producethese new story books Cheaply enoughso that theparents might purchase them at the end of the year for a nominal fee, two benefits might accrue. First, the reading skills acquired during the school yearwould not deteriorate quiteas rapidly over .the long semester breaks (4 months) because the children wouldhave at least something to read. Secondly, the parents might bestimulated to learn something from these books.

3.6.2.1.4ALPHABET BOOKS

In keeping with recommendation 6, the preparation ofanalphabet book was

57 PAGE 44

approved. The group felt that thiswas arelatively easytask andwould produce a greater familiarity with letters which, in turn, might improve the effectiveness of the letter drills.

3.6.2.2 Cycle 3 Innovations

3.6.2.2.1 FuNCTOR DRILLS

The functor drills turnedout to be the most controversialsubject at the workshop. There was considerable debateasto their value. Aboutthe only thing everyone agreed on wasthat selection of functors to be contrasted should be based on semanticcriteria. Cowan arguedfor a new format which provided for four sentences only one or two ofwhich Latched apicture and provided examples of hull thiswas done in commercial readers.When Wanda Pace agreed to try out this approach using Hdogo,she discovered two problems. Picturablecontrasts liketheone shown in Figure 5.1,worked fine and permitted the teaching of sore than on functor per lesson. The question was raised:what does one do when the contrasts arenot picturable? And the . answer "teach them without pictures" did not satisfythe group.

Another problem wasthe obvious impossibility of havingone procedure for bothcontrasts with pictures and thosewithout. Afterconsiderable discussion, it was 'decided toproceed with a modification of thepresent format proposed by Rick Brown. This isdescribed by himin detail with a rationale for all elements and proposedprocedures in Appendix 31!. 3.6.2.2.2 REVIEW

The review lesson was changodso thatall letters, words and functors selected for review would appesr in sentences.The rationalefor this change isrooted in recommendation3: thatallexercises should emphasize that reading is for 'waning. A planned match of pictures tosentences was devised by Julie Van Dyken. This should be effectivefor reviewing letters and words, but it will not permit the review ofgrammar points which are not picturable. Thiswas, however, not consideredto be a serious drawback. The format developed by Julie was to beconsidered tentative and mightundergo changes when tested in Cycle 3. An example.of thenew revision lesson and instructions for writing and teachinga lesson are found in Appendix 31.

3.6.2.2.3 UNDERDIFFERENTIATION DRILL

A new type of drill beingtried out is theunderdifferentiation drill, whichis supposed to provide the child with practice indifferentiating sounds. Cbwan Was given a sketch of what it looks like. The one thing that struck Cowan about the drill was thatit contained no provisionfor testing whether the children couldactually perceive Lhe differences thathad been practiced. He therefore suggested that some sentences whichcontained words that have the sounds be addedas atest and that a procedure be devisedwhere the teacher determines if the childrenhave indeed mastered the distinctions.

58 PAGE 45

3.6.2.2.4READING PASSAGES

Rick Brown suggested that atsome point the reading passagesshould have the questions which are normally asked orally printed on the same pageas the passage, so that the children may gain .practice readingquestions. This is in fact the approach taken in most comsercial English readers. However, due to tise limitations this proposalwas never discussed.

3.7 CYCLE 2 ORTHOGRAPHIES

Cowan was only able to examine the Kreshorthography in detail and get some feeling for the writing systemdevisedforNuer. Rick Brown, whowas supervising thepreparation of the Kresh material, arguedthat tone must be 'marked, since sany words are distinguishedby this feature alone. Cowan could not determine whether thefive tonelevels positedby RickBrownwere phonemic, but notedthat thepractice of marking alltones presented yet another learning task for a child. Not a greatdeal of data existson how much longer it takes to learna language where tone must be marked. The only case that comes to mind IS Thai 'and this is aslightly specious comparison, since that language has an overly complexsystem which usesconsonants and diacritics to indicate tone patterns.Cowan felt that Kresh would undoubtedly be harder to learnbecause thechildmust attend tomany diacritics in addition to the alphabet symbols.

The general SIL approach to devisingorthographies involves using diacritic marksrather than indirectsymbolization and this naturallycreates the additional learningprobles referred to. For example, insteadof signifying breathy vowels in Nuer with an h, underlining,a mark that is easily missed in reading, waschosen. Cowanfelt that a lecture series on the historyand principles of orthographyconstructionwould beusefulsince verylittle useful guidance is provided in the Gudchinsky(1973) Manual.

3.8 THE ARABIC MATERIALS

Rick Brown described in detail the oral/auralArabic program and the Arabic reading material he was developing for the primaryschools.

During the first year and a half that theChildren were learning to read in the vernacular, they were toreceiveinstructionin spoken Arabic. The rationale forthis was that the studentsmust learnto speak the language before theycan learnto read it. ButRick Brownalso claimedthat tests conducted in Juba (there was somevagueness about who had done the testing and what tests were used) had shown thatsany Sudanese have tl'ouble distinguishing sounds whicp havedistinctly differentrepresentations inArabic, aln/ and 06 /s/n/. It was assumedthat this would seriously affect their ability to read and hence the oral/auraltraining should address this problem.

59 PAGE 46

The oral/aural program Brown was devising spanned50 weeksand employed several kinds of drills: (a) total physicalresponse, (b) same-different discriminationdrills, (c) "oddball" drills (whichof the three orfour sounds is different?) and (d) production drills where the studentswere to be asked to make a sound which a certain aninalmakes, or some othersound in their environment similar to an Arabicletter. For example, they might be told to make a sound like "gargling" inpreparationforlearning pharyngeal t /yain/. the

Learning to speak Arabic was to begin withtotalphysical response drills where the children do what is saidin Arabic. Later, production drills would beintroduced as an extension of this. Childrenwould then respond to questions such as: "Wheredidyouput the pencil?" Conversational mos-playing and experience charts youldbe used togenerate conversation. Tts tftacher sight tell stories wi:tthe help ofthe chart and thenask qw.:Alons about the story or have thechildren retell the entirestory. It was intended thatthe children becomr familiar with the range of politeness forwulas which are mandatory in Arabiubut absent from or lacking equivalents in Sudanese languages.

The reading progras began with prereading work: sas different drills like those in the vernacular primers and entire wordsarrayed intriplets. The first lesson of the primer has two sightwords displayed next to two pictures, e.g. a name, John,and the word chicken% Below the pairs of words and pictures was a picture of John carryinga chicken.Onthe opposite page was the sentence: "John is carrying a chicken." The nextlesson distinguished the phrase "to the market" from the word"market."This contrast was set off in a box like a typical functor drill. The teacher wouldpoint out the difference between the phrase and the wordin terms of. the letters used.On the next page the sentence "John iscarrying the Chicken to the market" would appear. The lessons proceededthis way until a fewsight words have been taught. In Lesson 7, the distinction marketvs. the market, was taught; this introduces a single letter change which affectsthe shape of the word.

For the first ten lessons, therewas a high rate of repetition so that the sight words are firmly established. In Lesson 10, the firstvowel- letter sequence was introduced: a/de/ vs. /re/. Lessons 12 and 13were revision lessons and lesson 111 began the fliastreal letter contrast: .5 /di/ vs. _) /4/. This was doneusing a format verysimilar to the letter drills in the current vernacularmaterials.

Cowan felt that the Arabic programshowed considerable promise. Rick Brown obviously knew the language well and he.had produced a syllabus composed of often original and creative exerciseswhich were blendedwith conventional teaching techniques. Withcareful monitoring, the program wouldprobably be adequate to the task.Cowan had only two suggestions.First, he felt that it might be wise to teach the students all of the lettersbefore they begin the sight word lessons. Recognizing Arabic letters and how theyare joined to one another issomething thatdoes not comeeasy forsecond-language learners, even if they are children. Secondly, Cowan suggested that the spoken Arabic training should be extended to afull twoyears. There is someresearch (Cowan & Sarmad, 1976) which shows that in a situation somewhat comparable to Sudan, successful performance in learningto read a second language is related to the extent that the second language is spoken athome. Since Arabic will

60 PAGE 47 rarely, if ever, be spoken at home, it follows that asolid foundation must first be established at school.

61 PAGE 48

4.0IMPACT OF THE PROJECT

The project's impact on the development of literacy skillswas investigated by aseries ofthree evaluationsconducted fromNovember 1980 to November 1982. Although the original project proposalcalled for an evaluation of the effects ofthe project on the selfimageand behaviorpatterns toward development of the participants, it was decided toomit these aspects from the evaluation given the difficulty of developing reliable and validinstruments to measure these variables, especially for young children. Since it had been hypothesized that any beneficial affectiveconsequences ofthe project would be due to students' success in reading their mother tongue, it was felt that the best return on any evaluation effort would beobtained by determining just how well thechildren involved in the project were acquiring literacyskills in their language.

In addition to determining the effects of theproject on the development of literacy skills bycomparing theskillsof childrenparticipating in the project with coaparablechildren .whe were not, the evaluationwas planned to obtain descriptive information on theschools, teachers, curricula and pupils involved in theproject aswell asinformation on the actualuse of the project aaterials. It was hoped thatthis background and process information would provide a better understanding Of thecontext and impact of the project and provide a basis for suggesting changes toimprove the effectiveness of the project. The first impact evaluation wasconducted by Cziko over a four-week period in November and December 1980.This evaluation included both project a. d non-project (comparison) Bari and Lotuhopupils at Primary 2. The second impact evaluation took place a year later inNovember and Decemberof 1981 during which time COwan and Cziko collecteddata from Bari, Lotuho, Nuer and Zande schools. The third impact evaluationwas undertaken during October and November of 1982 by COwan and involvedPrimary 1 Zande project and comparison pupils and Primary 1 andPrimary 2 Muerproject pupils. Althougha fourth final impact-evaluation visitwas planned for the fall of1983, civil unrest in the region forced its cancellation.

This chapterdescribes thethree impactevaluations and discusses their results. The design of each evaluation is first presented, followed by the results which comprise background informationon the schools, teachers, school activities and the outcome of literacytests administered to both project and comparison primaryschool pupils. Thechapterconcludes with ageneral suamary and discussion of the impact evaluation findings.

62 PAGE 49

4.1FIRST IMPACT EVALUATION

4.1.1 DESCRIPTION OF EVALUATION

4.1.1.1Evaluation Design

The design of the first impact evaluation months as information evolved over a period often on the project and participatingand comparison schools was collected during three visits by Czikoto Southern Sudan inJanuary, June andNovember/Deceaber 1980. The principalfeature ofthe design comparing.two Bari and Lotuho involved schools, one school of eachpair using the new project materials (to bereferred to as project schools), referred to as comparison the other not (to be schools). Of the fourCycle 1 project materials were being languages for which used, Bari and Lotuhowerechosen due relative accessibility of to the rural schools using theselanguages regional capital. from Juba, the The four rural priaaryschools actually chosen (Rejaf Bari project; Bungu, Bari East, comparison; Hiyala,Lotuho project; and Loronyo, Lotuho comparison)were among the ten Bari and Lotuho schools visited inJune 1980 which provided atleast basic necessary educational facilities, 1,e., shelter, blackboards, chalk, paper,and writing noted that these four instruments. It should be schools should not be consideredin primary schools in rural any way typical of areas of theSouthern Sudan but rather since they appeared to were chosen provide whatcould be considereda reasonable chance for the literacy project tomake its impact felt.

Originally, it had been plannedto focus on the Primary fourschools in .November/Decemb0 1 pupils in these 1980 and then tofollowup these same children at the end of Primary2 in November/Deceaber 1981. discovered that the 1980 However, it was Primary 1 children at theBari project school had been able touse Book 2 of the not Bari aaterials since thePrimary 2 pupils at the sane schoolwere still using these materials ad there werenot enough copies for bothclasses. Therefore, it was decided tomodify the evaluation design by focusingon the Primary 2 children inbothBari and Lotuho schools inNovember/December 1980 and to follow up thesesame children in November/December 1981 at the endof Primary 3.

4.1.1.2 Instruments

To collectbackground information, questionnaireswere preparedand left with the Bariand Lotuho language officers in June 1980. Completion of the questionnairesrequired supplyinginformationon the (enrollaent at each grade), size of eachschool' the number of days insession during the 1980 school year, the number of Primary 6 pupils who passed theprimary school leaving exam at the end of the 1979 schoolyear, thecurriculum (subjects taught at each grade, bywhom, using what materials), teaching experience, the teachers (education, subjects taught) and thePrimary 1 and 2 pupils (name,

63 PAGE 50

mother tongue, age, and attendance). It was hoped thatthis background data would provide information on the context ofthe project and indicate theways in which the four schools differed.

The instruments developed for the collection of outcome datainvolved the creation of measures of theactualreading performance ofthe Primary 2 pupils. A group-adninistered reading test andindividual tests of oral reading and comprehension were constructed. The group-administered testswere adaptations ahd translations of the reading subtests ofthe Primary Level 1 Stanford Achievement Test. Tests of vocabulary, reading andword study skills were constructed in both Bari and Lotuho.The vocabulary testrequired the pupils toindicate the one word (out of three choices) which bestfit the sentence read by the examiner. The threewords were read by the examinerand the pupil had to. .indicate the correct word printed on his answer sheet.The reading test consisted of two parts.Part A involved indicating whichone of three printed words best described a picture presented above the words.Part B involved choosing the.one word(eabeddedin groups of three) which best a short printed "story" fit consisting of a sentenceorseveral sentences. Pictures were used to provide some contextfor each of the stories.Finally, the word study skills test was atype of word recognition task. The examiner read a word aloud twice and the pupilwas required to indicate the target word from a group of three words printed on his answer sheet. The correct word and the two distractors were usually similar visually and auditorily (anEnglish example would be stiff, stitch, stick).

The oril reading tests consisted of fourparts: (a) a listof ten words included in the project materials for Primary 1 and 2, (b)a list of ten words not contained in these materials, (c)a short story of approximately 50 words containing all of the wordsin the two lists and(d) five comprehension questions based on the story. These oral reading materialswere constructed in both Bari and Lotuho.

4.1.1.3 Data Collection

Three procedures were employedto collect thebackground, process,and outcome data. NOst of the background data were collectedby theBari and Lotuho language' officers between JulyandNovember 1980 viainterviews with school staff and referring to schoolrecords with theschooldirector. Additional information was obtainedduring the time of the author's visits to the schools in November/December 1980.While much of the information desired was collected, some was not available. For example, some of the age data that were obtained in most cases represent thebest guess of the teacher, pupilor language officer. Finally, since some of the teacherswere not present at the time of the evaluation visits, some of the information on teacher'seducation and teaching experiencemay not be completely accurate since it was obtained from secondary sources.

Process data were collectedon theactual teaching and learning activities of thePrimary 2 vernacular classrooms at eachof the four schools. These data were ottained by tape recordingand taking notes on a complete vernacular reading lesson which the author attended, assisted by the language officer of

6 4 PAGE 51

that language Who was fluent in English and was able toprovide with explanations and the author interpretations of the classactivities. The outcome data were obtained usingthe group administered andindividual reading tests describedabove.The group reading tests all Primary 2 children present were administered to at the four schools at thetime of_the author's visits in November/December 1980. The first group to beadministered group reading tests was the Primary the 2 class of the Bariproject school.Since these pupils hadgreat difficultyin understanding and vocabulary test (and completingthe in fact as a group performedonly at a chance level this test), on they were not administered the second and third PartsA and B) but proceeded tests (Reading: directly tothelast testof word Consequently, the word study skills. skills test, which appearedto be the easiest of the group tests, was the onlytest administered to the classes. A total of 83 Primary 2 other three Primary 2 pupilsin the priaary schoolswere administered this group readingtest. The oral reading test individually to 43 Primary was administered 2 pupilsin thefourschools. Although limitations prevented the time individual testing of allPrimary 2 pupils present, an effort was made to includea representative sample from Primary 2 classes each of the four including weak, strong andaverage pupils These save 43 pupils from each class. were interviewed to determine theirreasons for attending school and learning to readand write. The individual oral reading the interviews were conducted tests and by the Bari or the Lotuholanguage officer under the supervision of theauthor. Each pupil's oral reading to the comprehensive performance, answers questions and interviewresponses were tape recorded.

4.1.2 RESULTS

4.1.2.1 Background Data

General backgrounddata from the four participatingschools isshown in Table 4.1. Examination of this table reveals a numberof important characteristics of the fourschools. Both project schools of pupil enrollment are larger in terms and staff although theteacher-pupilratio was identical for the schools of eachlanguage group (1:36 for the Bari the Lotuho schools). schools; 1:21 for Both project schoolsseea more successful in preparing their pupils for the primary school leaving examination, the is prerequisite to. entering passing of which junior secondary school. Amongthe Bari project school and Lotuho projectschool Primary 6 pupils who January 1980, 32% took this examination in and 52% passed, respectively,while these and 12% for the Bari comparison figures were 16% and Lotuho comparisonschools, respectively. With respect to teacher educvtion, the twoproject schools had proportion of teachers who a higher had done atleast some studies ata training college either in teacher Sudan or Uganda (75% and 60%for the Bari project andLotuho projectschools, respectively, vs. 40% and 43%for the Bari comparison andLotuho comparison schools, respectively). The Bariproject schoolalso differedmarkedly from the otherthreeschools in termsof teachers' work experiencewith the median number of years of primaryschool

65 Table 4.1

Background Data

School

Rejaf East Bungu Riyala Variable Loronyo (Bari Project) (Bari Comparison) (Lotuho Project) (Lotuho Comparison) alMff1=all.PIMNBM.M. School Data Pupils enrolled in 1980:

P1 79 38 40 31

P2 82 24 35 14

P? 44 13 27 19

P4 47 32 28 18

P5 14 40 29 20

P6 26 33 48 43 Total 292. 180 207 145 Teacher/Pupil ratio 1:36 1:36 1:21 1:21 P6 pupils in 1979 53 56 31 32 Above pupils passing

leaving exam 17(31%) 9(16%) 16 (52%) 4(12%) First day of class in 1980 8127 8/25 5114 6115

(coned)

66

67 talk 111 Iterill

kkenvo1 Dali

MINUE1111111111111111111111111110111111111

army

4110 IAN Soy Illyels Loony° Moo 14N11 1111,1 Comprissn1(Let*. Prolect) (lotuho Comparison) Coot hi* OM

law 0 istrs I i lo

Om INOw WM A %lag nap Wipe 1 WO 1 Mil 7 (711) 1 (NI) NNWloong

ilio It w II III

bp Ia 11 1 a 12 NNW WOO i PM 1 11111 I 1M1 3 1131) .111110111WIWI OW N 1 I i lop I s Il 1/0.1 713 - 27 2 - 21

(cont'd)

69 Table 4.1 (cont'd)

Background Data

School

Rejaf East Bungu Variable Hiyala Loronyo (Bari Project) (Bari ,Comparison) (Lotuho Project) (Lotuho Comparison)

Primary 2 vernacularteacher information

Mother tongue Education (years) 7 8 8 6 Training

TTCd yes no yes no SILe yes no yes no Experiencet(years) 35 6 27 5 Experience: P1 6 P2 (years) 15 6 27 5

Primary 2 pupil information

Enrollment:

Total 82 24 35 14

Male 47(57%) 16 (67%) 32 (91%) 14 (100%) Females 35 (43%) 8 (33%) 3 (9%) 0 (0%) Regional

mother tonguea 72(88%) 24 (100%) 34 (97%) 14 (100%) .

(cont'd)

70 Table 4.1 (coned)

Background Data

School

Re Of East Bungu Variable Hlyala Loronyo (Bari Project) (Bari Comparison) (lotuho Project) (Lotuho Comparison)

Primary 2 pupilinformation

Age:

Mean 9.38 9.79 10.42 12.08 Standard deviation 1.08 1.25 1.90 2,25

nf 70 24 7 13

aRegional mother tongue refersto Bari in Bari schools and Lotuho inLotuho schools. b Trained teachers are defined as teacherswith at least some teacher training collegeattendance. c Experience refers to years of working asa primary school teacher, d A yes. in thisrow indicates some attendance at a teacher trainingcollege. e A yes. In this row indicates participatingin the SIL teacher training foruse of the project materials. fNumbers in this row referto the number of pupils for whomage estimates were obtained.

72

73 PAGE 52

teaching experience equal to 30 for the Bari project schoolcompared to 5, 8 and 6for the Bari control, Lottiho control and Lotuho comparison schools, respectively. This is likely due to the fact that the Bariproject school (Rejaf East) is the oldest school ofthe four, having beenestablished by Italian missionaries after World WarI.

The Project 2 vernacular teachers at thefour schoolsdiffered markedly with respect to education, training and teaChing experience. In general, the project teachers had received training fromboth a teacher training college andfrom SIL on the use of theproject materials while thePrimary 2 vernacular teachers at the two comparisonschools had had no teacher training of any kind. Also, the two project teachershad from2.5 to 5 times more teaching experience at Primary 1 and 2 thandid thecomparisonschool teachers. This finding reflects the factthat the project sought experienced teachers who had taught vernacular literacy skills in missionaryschools to participate in the first trial of thenew project materials.

Therewere large differences in the size of Primary2 enrollment, the percentage of female pupils and the estimated age of thePrimary 2 pupils in the fourschools. Both project schools had markedly more pupils than their corresponding comparison schools. Also, while females madeup a sizeable part of the Primary 2 classes at the two Barischools, therewere onlythree females in the Primary 2 esss ofthe Lotuho project school andno female pupils in the Primary 2 class of the LotUho comparison school.Almost all of thePrimary 2pupilsin the Bari comparison, Lotuhoproject andLotuho comparison schools had the regional language as their mother tongue (Bari or LotUho) while10 out of the 82 Primary 2 pupils did nothave Barias their aother tongue in the Bari project school. These 10 pupils were refugees from Zaire whose familieshadrecently settled nearRejaf East. The mean approximate age of the Primary 2 pupilsin each of thefour schools was generally comparable except for the LotUhocontrol schoolwhose Primary 2 ptIr'is were generally older with an approximate mean age of 12 years. Since cliled. supnosed to start Primary 1at age seven and therefore Primary 2 at age eight, the ages of the fourgroups of Primary2 pupils indicates that it isnot unusual for children start their primaryeducation at 8, 9or 10 !ears of age.

Inforiation on the Primary 2 curriculum :Admaterialsused in the four schools 21 presented in Table 4.2.The most striking aspect of the and matials used curriculum in thefour Primary2 classesis the serious lack of materiP.sfor the pupils. Ofthe 20 subjectby school combinations !lore-antedby thefour Primary 2 classes, onlyfive of theseincluded for thepupils. Noneof the 'ourPrimary 2 classes hadpupil skserials for mathematics and for three of theschools, only one Primary 2 subject wzs taught with the aid of pupil materials. Of particular relevance to this evaluation is the fact that wiLile the Bari project 07.,IT had the new project aaterials for about half of the PrimAry2 pupils in the vernacular class, all pupils in the Bari coeparison Prfziary 2 clamhad for class use a copy of a Bari primer first published in 1958by L. J. Fallon, Ltd. (Dublin) and republished by the Regional Ministry of Education (Juba) in 1976. Unlike the Bari project teacher, however, che Bari comparison vernacular teacherdid not have a teacher's manual fOr this primer nor training in the use of the primer. For the too Lotuho schools, the Lotuhoprimary school wasable to supply all of its Primary 2 pupils with copies of the new project materialsto

7 4 Table 4.2

Primary 2 Curriculaand Materials

Teacher's Pupil's Minutes/ Materials/ Materials/ % Pupils Subject Week Medium 1111111.,MMINNOMIN Language Language with Materials

Rejaf East (Bari ProjectSchool)

Vernacular 240 Bari project/ project/ 46% Bari Bari

Mathematics 240 Bari RMEI none 0% English

Religion 80 Bari Catechism/ Catechism/ 100% Bari Bari

Bungu (Bari ComparisonSchool)

Vernacular 270 Bari old/ old/ 100% Bari Barl

MatherrItics 270 Bari RMEI none 0% English

Religion 270 Bari Catechism/ none 01 Lotuho

ti ygiene 270 Bari Yuggu/ none 0% Bari

(cont'd)

75

76 Table 4,2 (coned)

Primary I Curricula andMaterials

Teacher's Pupil's Minutes/ Materials/ Subject Materials/ % Pupils Week Medium Language Language with Materials 41MINNIPMEMINiWINNIIWIM11.6=~1FRI Hiyala (Lotuho ProjectSchool) .01~1MiaMMIENM.M.D10,1PVIN..e.

Vernacular 200 Lotuho project/ project/ 100% Lotuho Lotuho

Mathematics 240 Eng lis' 6 RME/ none 0% Lotuho English

Religion 160 Lotuho none none 0%

English 240 English 6 OUP1 none 0% Lotuho English

Arts 6 Crafts 80 Lotuho none none 01

Loronyo (LotuhoComparison School)

Vernacular 240 Lotuho old/ none 0% Lotuho

Mathematics 240 English RMEI none 0% English

(coned)

77 73 Table 4,2 (ccnt'd)

Primary 2 Curricula andMaterials

IPIMMIMMEIMIIIMMIIMMIMININIMMIM Teacher's Pupil's Minutes/ Materials/ Ma terials 1 Subject % Pupils Week Medium Language Language with Materials

Loronyo (Lotuho ComparisonSchool)

Religion 240 Lotuho Catechism/ none 0% Lotuho

English 240 English Oxford/ Oxford/ 100% English English

Hygiene 80 Lotuho none none 01 Arts 6 Crafts , 80 Lotuho none none 0%

Arabic 240 Arabic CME/ none 01) Arabic

Physical Education 80 English none none 0%

Note.Abbreviations used to specify materials referto their source, i.e., RME = RegionalMinistry of Education (Juba); CME= Central Ministry of Education (Khartoum); OUP= Oxford University Press,

79

00 PAGE 53

use duringthe vernacular class whilethe teacher used the teacher manual developedfor these materials. Incontrast, thePrimary 2 pupils atthe Lotuho comparison school had no educational aaterials in Lotuho. The only materials available to the Primary 2vernacular teacher at the comparison school was an old copy of the Lotuho primer previouslyused to teach reading in Lotuho. It vas also brought to the author'sattention that this battered copy of the primer was only located a few days previous to the author'svisit to the school on 1 December 1980 and consequently its use with the Primary2 pupils at this school WUvery recent.

Although as mentioned above none ofthe schools kept complete attendance records, it is important to make a few comments on the number of days thefour schools were in session in 1980. Whileprimary schools in the SouthernSudan are supposed to open officially in April and close just before Christmas,all four schoolt began sessions considerably later than April in 1980. The Lotuho project school began classes on Nay 14 with the Lotuho comparison schoolabout a month _later on June 15. The Late opening of thetwo Lotuho schools seemed to be due primarily to the holding of regional elections in Apriland May 1980 for which priaary schools were used as polling places.The opening of the two Bari schools was further delayed by ateacher strike andit wasnot until Augusty 25 and August 27 that the Bari comparison and projectschools began, .respectively. In addition, at the tine ofthe author's evaluationvisit in November 1980, the four schools were closed for one week of "localleave" with relatively few Bari pupils in attendance oneaeek later, therehaving been some confusion over whether local leave wasto last one or two weeks. This unexpected break seriously affected the mount of time the authorhad to visit these schools and resulted in testing fewer individual pupils thanoriginally planned and the elimination of the elicited imitation task fromthe evaluation design.

Finally, it should be mentionedthatthe Primary 2 project pupils included in this evaluation began using the new project materials in Bariand Lotuho October 1979.Since this was already towardthe end of their Primary 1 school year and since the beginning of theirPrimary2 yearwas delayed, their performance on literacy tasks couldbe expected to be poorer than if they had begun use of the aaterials at the beginning of Prinary 1 and had theadvantage of school years of normal length.

To summarize the background informationcollected from the four schools, it should be quite clear that the two comparison schools should notbe considered control schools in any realsense of the word. Thebackgroun6 Information reveals systematic differences betweenthe twoproject ,schoolt, and their respective comparison schoolson aany important factors.These faors, which appear to favor the two project schools, include enrollment, proportion of primary school leavers qualifyingfor entrance into junior secondary schools, number of teachers, teacher training andexperience and theavailability of pupil materials. In spite of these differences,however, the inclusion of the two comparison schools nevertheless provide some baseline information on the development of literacy skills in ruralprimaryschools which had not had the intervention of the literacy project.

81 PAGE SI

ea 4.11

senticome dials obtalase fOse the unservelleas andtape recordings lat ainnInalo Ohs Oarl prejesto Sari semperissa, toteho projectand SIONFINI ire sonnmelmod and presented inTibias 4.3, 4.4, 4.5 and tassInnit lbws IOWInver= talleztlastaislyzre er classroom mom oo0 prom No:ALla Ohs order is shish the;took place 16 a. olegopmee. Ulla It le POOMIMASSOthat Ohio sieplitiod scheme Is Ls paw alObi elle Malls ef OlOOSPOSO estivate',it is telt 111"==:. OM Muni inforeatiaa en Ohemajor teaching and IMO aft plods la theMoor Glasses. This is due to the gni&*UM/ all olouross'Militia, Involved the MOON, Giro SW Ols0 Ohm teacher and the sere very few latersetleas between teachers and Salomoh leamhors and mall MIAS Olorgoos. groups of pupils, or among

So fedls;-8 Ohe osgeombort deeeriptleas of class activities NO1100010 On ) lirmagallol as is Week by the oempletelack ot oNIMPUlooftUs Om somporliolo dos 0.taNprawnmums as a anninsw So oftempt moproonOoattheyread. &Mot* US le NO Smoke am massingthlremain aotivities (Table 414, %Web Lilliesos. 5, U),loon oi~el activitiestook to 11111Merl Pri eds..,me towbar radthentIllurritordralots: ElLelt#46.arLlialtagoolft=selroulng gosatlieteleledi OIL Itelaserptiparli; Ihinforo, ta OW dos Itas pissill thatreading sooproloalloofor so or sot lowolvW to ansusrlastoequestions based slaftp. on the

loo NNWIssisii Ohms, et Ohs vernasslar readiesactivities in the fur alooloI. ostairities senslat 'ghost entirely otgroupreading sera ad unrelated seatemoes withsome practice "Is =rose ev=viOral inros. Ill000loomUssoosOootObsom CI four Glasses uhlob do netappear in Tables Ohm. keoroompoo mumto loth preps% elesseseagreatdeal of Op She lomeber In writing syllables,words sad sentences on the sop SOO aol ofenlerlals written 16.11-1711111re ea the hoardMoot amOsilale6 Sorbs Obese times, the tempoof the oleos ouplolsOp abllo Ohs lusher eery slowly(although very clearly) 0111411111 amens Moly that the projectteachers were MAU sillelmmilineOn Oho pupils Waft theirewe materials andfelt more smorsimie 111marstAimgtheir resalesastivitiesby plating to parts of as Ilisrlerre. She classes the repetitica of soundsand eons ems ammo idle end ellen leek en a rhythmic, cheating quality WM mewl OS able arlsie Mope ef She elesereem activities. In one elmPori ouporlamil, a.4emeher mewl. got the pupils Lorepeat the word Op Mop* amplimg sags,shish NNW often be dem severaltimes in quick momemeloa. Ibisla subtlisasilbe underseas the finding thtt thesounds of Ur eftsamilWOMBla *Loh they were repeatedsees to be given much

82 Table 1.3

!.4utntial Description of VernacularClass Activities; Rejaf East (BariProject Class)

No. Stimuli Respondent Response

1 sentence on board class reads sentencealoud 2 nonsense syllables on board class reads syllablesaloud 3 sentence, phrae, wordon board class reads board displayaloud 4 sentences on board Ciass reads sentencesaloud syllables un board class reads syllablesaloud 6 syllables, sentenceson board individuals points to syllables, sentences reads themaloud function on board class reads functionaloud 8 teacher reading aloudunrelated sentences from manual class listens to teacher

9 teacher's instructions Individuals produces sentencesfollowing grammatical pattern(not connecteddiscourse, sentences, not writtenon board) 10 sentences on board class reads board aloud

11 words, sentenoson board, teacher's reading aloud class repeats words,sentences after teacher 12 syllables, words,sentences on board Individuals reads board aloud

13 teacher's pointingto functions in sentences on board class reads functionsaloud 14 Book 2,p. 60 (letters, syllables words) Individuals readpage aloud

(coned) 83 84 Table 4.3 (coned)

Sequential Descriptionof Vernacular Class Activities: Rejaf East(Bari ProjectClass)

No. Stimuli Respondent Response

15 teacher's pronunciationof Lit, class repeat teacher's pronunciationofLai 16 story on board teacher reads board aloud

17 story on board class reads board aloud

18 story on Book 2,P. 61 instructions to read silently class reads page independently, but notvery silently 19 repeated instructionsto read silently class reads 'more silently"

20 story in Bo* 2,P. 61 teacher reads story aloud

21 story in Book 2,p. 61 class reads story aloud

22 story in Book 2,p. 61 all girls reads story aloud

23 story in Book 2,p. 61 all boys reads story aloud

24 teacher's questionson story individuals answer one question

25 11., it it! on board,eight times each Individuals reads board aloud

26 lit it, MIon board, eight times each class prints words in exercise book :orrectswork 27 teacher's dictation of1.01k,koti class prints words inexerck4 book 28 student's written workfrom nos. 26, 27 teacher corrects work

85

86 Table 4,4

Sequential Description ofVernacular Class Activities: Bungu (Bari ComparisonClass)

No, Stimuli Respondent Response

1 story In book; teacher'sreading class reads (repeats) each of each sentenc.? sentence aloud after teacher

2 story in book, teaCr''s corrections class reads story aloud 6 guidance

3 words on board class reads board aloud

4 words on board dividedinti i:bles class reads syllables aloud6 blends Into words stc In book, teacher'sguidat., tt, 6 class lions reads sentences aloud

6 story 'It; each senten 4:1=s5 reads sentencesaloud aloud tlacher

story in rr, individuals read story aloud

8 word prontmced slowlyby teather Individuals print wordon board (different word for each pupil)

9 test heading on board class copies of test headingon test paper (name, date, "Bari Spelling")

10 25 words pronouncedseveral times class prints wordson test paper slowly by teacher

11 25 spelling wordspronounced quickly class checks spelling by teacher on test paper

87

88 Table 4.5

Sequential Descriptionof Vernacular Class Activities: Hiyala (lotuhoProject Class)

No. Stimuli Respondent Response

1 picture of antelope withabobito class says abobito on poster

2 list of wordson board class reads words aloud

3 list of wordson board individuals points to and readsaloud all abobltoin list 4 words, syllableson board class read aloud

5 teacher's instructions individuals say words beginning witha given letter 6 sentences on board individuals say new sentences beginningwith word ofeach given sentence

7 Book 2, p.56 (words,syllables, individuals read page aloud letters)

8 words on board, teacher'sinstruc- Individuals say new sentences containing Hons., teacher's sayingof word given word

9 story in book, teacher's instructions Individuals read story independently,silently ,, 10 teacher's questionson story (new individuals answer one question question for each pupil)

11 teacher's reading storyaloud class listens to teachers

12 story in Book 2,p. 57 individuals read story aloud

13 words on board class prints words in exercise book (copywords letter by letter)

(cont'd)

89 90 Table 4.5 (cont'd)

Sequential Description ofVernacular Class Activities: 'Ilya la (Lotuho ProjectClass)

No. Stimuli ,111111111.111 Respondent Response

14 sentence dictated by teacherword class prints words in exercisebook by word (each wordrepeated four times)

15 student's written workfrom no, 14 teacher corrects written work

04c40 91 Table 4.6

Sequential Description of Vernacular Class Activities; Loronyo (LotuhoComparison Class) ..... No. Stimuli . Respondent Response

1 nonsense syllables on board read class read syllables aloud afterteacher aloud by teacher

2 words on board read aloud by class read words aloud afterteacher teacher

3 words on board class reads words aloud

4 sound of n given by teacher class repeats sound

5 syllables, words 6 sentences on Individuals read board aloud board

6 syllables, words 6 sentenceson class board copied in exercise books(done at board author's request)

93

94 PAGE 55

sore importance thanthe meaning of the words andthe communicative function of literacy.

Finally, the analysis of the errors made by soae of the Bari projectpupils in answering the class comprehensionquestionsis revealing. Since many of the comprehension questions can be answered by simply repeatingan appropriate sentence or part of a sentence from the story, it was often not clearwhether the pupils were actually understanding the stories or simply memorizingthem from repeatedly hearing then read aloudby theteacher and el! the class. However, three pupils duringthe Bariproject literacy lesson began answers to comprehension their question with the worda, which means and and then and is often used at the beginning of a non-initial sentenceof a Bari story for the purpose of text cohesion. However,answering an oral question witha sentence beginning with a is not appropriate (in factthe teacher vigorously corrected these pupils) and seems to indicatethat these pupils had in fact memorized the story and went so far as to violatesome basic discourse rules of spoken Bari to use what they had memorized to answer thequestion.

4.1.2.3 Outcome Data

The outcome data of the first impact evaluation of the literacy projectare summarized in Tables 4.7 and 4.8. Before examining theseresults, however, several factors should be kept in mind. First, a relatively small numberof pupils were adainistered the individual readingtests. Usaally, this was due to low attendance at the four schools and/or lack of time. Therefore although the pupils tested are believed to constitute a fairly representativesample of all of the pupils in each of the four classes, they were not randomlyselected from theirPrimary 2 class. Also, there were important differencesbetween the two comparison classes. While the Baricomparison classwas able to provide all of its Primary 2 pupils with reading material in the vernacular, no aaterials whatsoever were provided for the Primary 2 pupils in the Lotuho comparison class. Finally, it should be kept in mindthatalthough tests of statistical significance (i.e., t-tests)werecomputed to compare the reading performance of each project class withits comparison class, these testswere performed for descriptive purposes only. Since the pupilswere not randomly selected from theirrespective populations, the results of these analyses should beconsidered to provideinforaation onlyon thedifferences in performance between the actual pupilsbeingcompared.Generalizations of these results are only valid insofaras these pupils, teachers, saterials and contexts are similar to thoseelsewhere in the Southern SuL:41,

4.1.2.3.1READING.PERFORHANCE

The reading test performance ofthese Bari and Lotuho pupils is summarized in Table 4.7. On List 1 (reading aloud 10 words taken fromBooks 1 and 2 of theproject materials)all groupswere able to readapproximately three-quarters of theft words incept the Lotuho comparisongroup which could read virtually none. On List 2 (ten words notappearing in Books 1 and 2 of the project materials), performance was, as expected, poorer.However, while the Lotuho project group performedmarkedlybetter than the Lotuho comparison group, the Bari comparison group performedat a higher level than did the Bari

95 Table 4.7

Reading Test Results

AINMIamMINEMII=11 School

Rejaf East Bungu Variable Hiyala Loronyo (Bari Project) (Bari Comparison) (Lotuho Project) (Lotuho Comparison)

List 1

14 7,6 7.5 7.7 SD 0.1 1.8 2,4 2.1 0.3 r 14 6 9 9

2,9 5.2 3.8 0.0 3,0 12 4.4 0.0 14 6 9 9 Story M 24.6 20.2 17.8 0.1 Ko 9.4 16.3 9.2 R.- 0.3 14 6 9 9

Comprehension Questions M 3.4 2.7 1.6 0.0 i-D 1.0 2.2 1,1 N 0.0 14 6 9 9 Word Study Skills

NI 32.2 52.3 42.6 23.2 -5-0 11.8 3,7 14.4 N 3,1 22 14 24 6

Note: Maximum score is 10 forboth Lists 1 and 2, 5 for Comprehension Questionsand 60 for Word Study Skills. Maximum socre for the Bari Story is 41 while thatfor the Lotuho Story is 52.A solid line between twomeans indicates a significantdifference (2 < .05)

96

97 PAGE 56

project group. The better performance of the Bari comparisonpupils on List 2 seemed to be dueprimarily to their ability touse lettaer names to sound out unfamiliar words. Almost all the Bari comparisom pupas attacked words by first reading the letter names ofthe component Betters and repeating these names with increasing speed until the word1413 recognised. No pupils from the other three classes employed this strategysince apparently they had not been taught the naaes of the letters orthe individual momndm that eachletter represented.

It WIW only while reading aloud thestory and mmowebing the comprehension questions that both project groups performedbeaker than their respective comparison groups. However, since it waspossibbe ffirethe pupils to answer these comprehension questions correctlywithout toilteoderstanding thestory or even parts of the story, it is likely that withItho exception of the Lotuho comparison pupils these scores overestimatethe pupitml abilityto read with comprehension. (The easy nature of the comprehensilen questions Is due tothe fact that they were constructed to be similar bothe comprehension questions used in the project materials.See the secOndpoint in thediscussion section.)

Finally,performance on thegroup-adminiZteredword studyskills test indicatedthat the Lotuho project group- ppeformed betterthan theLotuho comparison group whilethe Bari comparisoolpsup performedata higher level than the Bari project group.

4.1.2.3.2 INTERVIEW DATA

The reasons obtained from tbe interviewedpupils forattending school are categorized and presented inTable 4.8. Themostpopularreasons for attending school were related toobtaining employment (61%of allpupils interviewed). The jobs specifically mentionedwere teacher (26%), government official (13%), doctor (5%) and nurse (3%). The second most popular reason given for attending school wasrelated tosome aspect of literacy (i.e., learning toread and/orwrite) with 42% of all pupilshaving given such a reason. The most popular remainingreasons weregeneral academicreasons (e.g., to obtain knowledge, to be educated)and reasons related to obtaining prestige in some way (26% of pupilsinterviewed gave oneor bothof these types of reasons). The prestige reasons included beccminga "big man" (16%), being clever (11%) and obtaininga car (3%).

With respectto group differences inreasonsfor school attendance, the Lotuho project is characterized by a relatively highproportion ofpupils having given general acadesic reasons andno pupils having given literacy-related reasons for school attendance. The Bari projectclass was characterized by a relatively low proportionof pupils having given prestige- and employment-related school attendancereasons. However, theBari project pupils were interviewed immediately afterthe administration of the individual reading tests while for the otherthree classes the interviewspreceded the administration of the individual readingtests. This may then explain why the Bari project pupils gave a high proportion of literacy-relatedreasons and a relatively low proportionof employment-andprestige-related reasons for attending school.

98 Table 4.8

Reasons for AttendingSchool

School

Rejaf East Bungu Hiyala Reason Loronyo (Bari Project) (BarlComparison) (Lotuho Project)(Lotuho Comparison) Total

General academic I(7%) 2 (33%) 6 (75%) 1 (11%) 10 (26%) Learn vernacular 3 (20%) 0 0 0 3 (8%) Be literate 8 (53%) 3 (501) 0 5 (56%) 16 (42%) Learn to read 6 (40%) 3 (50%) 0 2(22%) 11 (29%)

Correspondence 3 (20%) 0 0 0 3 (8%)

Books I (7%) 0 0 0 1 (3%)

Learn to write 5 (33%) 0 0 3 (33%) 4(21%)

Correspondence 3(20%) 0 0 0 3 (8%)

Learn mathematics 2 (13%) 0 0 0 2 (5%)

Obtain prestige q (67%) 3 (38%) 3 (33%) 10 (26% Be a big man 0 1(17%) 2 (25%) 3 (33%) 6 (16%)

Be clever 0 3 (SO%) 1(12%) 0 4 (11%)

Have a car 0 0 1 (12%) 0 1(3%)

(confd)

99

I 0 Table 4.8 (cont'd)

Reasons for Attending School 1,=.....MINNIMI,...~1MM.=.11.Nmx..1 School 11.IMOMINI1=111=11111111111MIMMINIMP.M.1110.1111M1Wg., Rejaf East Bungu Hiyala Lorenyo Reason (Bari Project) (Bari Comparison) (Lotuho Project)(Lotuho Comparison) Total

Find employment 3 (20%) 5 (83%) 7 (88%) 8 (133%) 23 (61%)

Be teacher 2 (13%) 2 (33%) 2 (25%) 4 (441) 10 (26%) Be nurse 0 1 (17%) 1 (3%) Be doctor 0 0 2 (25%) 0 2 (5%)

Be government official 0 2 (33%) 0 3 (33%) 5 (13%) Other 1 (n) 0 0 1 (3%)

Learn Arabic 1 (71) 0 0 1 (3%)

Number interviewed 15 6 8 9 38

Note:Indentation ofreasons indicates subordination. For example, 8 Bari Projectpupils gavea "be literate" reason for attending school of which6 pupils specifically gave a "learn to read" reason.Of these latter 6 pupils, 3 further specifieda "learn to read correspondence" reason.Since superordinatereasons include all pupils givinga more specific subordinate reason plus less specific pupils and sincepupils often gave more than one reason, sums of columns percentages exceed100%.

101 1n2 PAGE 57

4.2 SEC0ND IMPACT EVALUATION

4.2.1DESCRIPTION OF EVALUATION

4.2.1.1 Evaluation Design

The second impact evaluation was originallyintended to follow up thesame Bari and Lotuho pupils btudied inthe first evaluation. However, since some modifications in the design of the literacy materials had taken place forthe Cycle 2 languages (due to the workshop held by Cowan in Maridi in October1980 described in Chapter3), it was feltthat itwas imperative tocollect information on the impact that these modified materialswere having on cne acquisition of literacy skills. Therefore, Primary 1 projectclasses using trial editions of the materialsprepared fortwo Cycle 2 languages (Zande and Nuer) were also visited, observed, and tested. Although itWas not possible to collect similar data from Nuerand Zande comparison classesat this time, it was felt that the data collected from the Nuer and Zande projectclasses'in itself would provide valuable informationonthe effectiveness of the Cycle 2 modifications and considerably broadenthe base of our evaluation. In addition tothe six classes mentioned above,intonation was also collected from two additional Bari and Lotuho classes not included In the first impact evaluation. Therefore, data werecollected from a total of eightclasses. The four Primary 3 classes followedup from the first evaluationwere Rejaf East (Bari project), Bungu (Bari comparison),Hiyala (Lotuhoproject), and Loronyo (Lotuho comparison) Primary Sthools. The two Cycle 2 classes wen the Primary 1 classes of primary schoolsinNasir Vest (Nuer) and Yambio (Zande). The two additional Cycle 1 classes consisted of a Primary 3 project class in Rejaf Vest (Bari) and the Primary 2 project class of Torit East PrimarySchool (Lotuho). Information was collected toward theend of the schoolyear in November and December 1981 by Mika (Bariand Lotuho classes) and Cowan (Nuer and Zande classes).

Because of timelimitations, we wereunable to visitRaga, where testing of the the third Cycle2 language (Xresh) was taking place. Jocelyn Clevengerreportedthat the IRL languageofficer and the two literacy inspectors who wrote the first threeprimers with SIL consultant Rick Brown were testing Books 1 and 2. Because of a teachers'strike, they were unable to begin before August 1981 and hence had only begun Kresh Book 2when our visit was made in November. SILconsultant JulieVan Dykenhad made one supervisory visit of about 15 days. The students had been divided groups on the into twa basis of language ability,onebeingcomprised of native speakers and the other of non-native speakers. The two literacy officerswere teaching the respectivegroups under thesupervision of the language officer. Regrettably, the onlypersonwho mighthavesuppliedus withdetailed information about plansfor expanding fresh instructionin other schools around Raga in 1982, Julie Van Dyken, was in Nigeria at the time ofour visit.

1 03 PAGE 58

4.2.1.2 Instruments

To collect background information, the same questionnaire thatwas used for the first impact evaluation was again usedto interviewthe director and teachers at each school. For thefour Bariand Lotuhoclasses thatwere included inthe first evaluation, onlythat information which hadchanged since last year was recorded (e.g., new teachers, class enrollments, Frimary3 curriculum and materials).

The instruments developed for the collection of outcomedata involved the creation of measures of thereading ability of both projectand comparison pupils in their respective mother tongue. For pupils in Primary 2or 3, a writingtest was also developed andadministered. The literacy tests developed for each of the classes are described belowfor each of the four languages included in the evaluation.

4.2.1.2.1 BARI TESTS

The Bari literacy test battery was made up of threeprincipal parts. The first part MS a test of word skills composed of four subtests. The subtests were (a) Vowel Words, (b) Consonant Words, (c) Known Words sad (d) Unknown Words. The subtest involved reading aloud 10 two-syllablecommon Bari words containingthe 10 vowels ofthe Barilanguage. Since orly sixletters (graphemes) are used to represent these 10 vowels (the tense-lax distinction not being indicated for four of thevowels), this subtest requredthe pupils ,o apply their knowledge of Bari vocabulary to decode each word. For each of the ten items of Vowel Words subtest, the vowel had to be read inone way only (either tenseor lax) to make a Bari word and the word had to be correctly readfor thepupil to pass the item. One-syllablenonsense words, each containing one of the 19 consonants ofBerl, lade up thesubtest. For this subtest, the pupilwas scoredon whetheror not he read theconsonant correctly, ignoring what was read for the vowel sound. The subtest consisted of 10 words used In Books 2 and 3 of the neu 3ari literacy materials whilethe subtest consisted of 10 words not containedinBooks 2 or 3 of the materials. The latter two subtests were the same ones that had been usedfor the first evaluation in 1980 and pupils were required to readaloud theentire word correctly. These tour subtests therefore comprised49 items.

The second principal part of thebatteryinvolved reading aloud a story of 52 words and answering six comprehension questions based on the story. These questions included some which referred to information stated explicitly inthe textof thestory as well as otherswhichrequired the inferring of information Implicitly included in thetext. Thelast part of the battery involved a dictation ofa short passageconsisting of atotal of three sentences and 16 words. These words were 511 included in Books 2and 3 of the new Bari literacy materials.

4.2.1.2.2 LOTUHO TESTS

The Lotuho literacy test batterywas also made up of three principal parts. The test of word skills comprised four subtests, viz., (a) Vowel Words, (b) Consonant Words, (c) Known Words and (d) Unknown Words.TheVowel Words subtestconsisted of five nonsense words eachcontaining oneof thefive

104 PAGE 59

vowels used in the Lotuho language. The Consonant Words subtest contained27 nonsense words, each containing one of the 27consonants of Lotuho. The Known Words subtest included 10 Lotuho words from Books2 and 3 ofthe new Lotuho literacy materials while the Unknown Wordssubtest included 10 wordsnot contained in these same materials.

The second principal part of theLotuho literacytest batteryinvolved reading aloud a story of 52 words and answeringeight comprehension questions basedon thestory. As for theBari test, thesequestions referred to info:illation both explicitly and implicitlystated in the reading passage.

The lastpart of the Lotuhobattery involvedthe dictationof a short passage consisting of a total of threesentences and 20 words. These words were all included in Books 2 and 3 of the Lotuholiteracy materials. 4.2.1.2.3 NUER TESTS

Both the Muer and Zande literacy tests were developedin the field during Cowan's visit. They were modeled after the Bari and Lotuho test batteries. The first subtest consisted of 12 nonsense syllableswhich contained letters the students had been exposed to in theprimer. Thepurpose of this subtest was to gain a general measure of the students' decoding ability. The second subtest comprised a list of 20 words, 10 of which were"known words", i.e., they were drawnfrom the lessons thestudents had studied upto the testing dates (Lesson 45 for the Nuer primer, Lesson 49 for theZande primer). The other 10 words were "built" words, i.e.,containing lettersthe students had been taught, but which had not actually beenused inprevious lessons. (The word bok, "book",.was used in both tests and although it is never used in the primer, it does appear on the title page.) The final subtest consisted of a short passage in which all of theprevious20 test words appeared. The original Nuer passage, which was 56 wordslong, had to be drastically revised after testing revealed it was much too difficultfor the pupils. The revision was then pretested, usingthree adults and a literate sixthgrader before continuing the testing. The passage wasfollowed by sixcomprehension questions; the answers to two ofthesewere notdirectlystated in the passage, but could be inferred from the content. The children read all of the various subtests aloud in the order described above (nonsensewords, 20 test words, comprehensionpassage) and werethen asked thesixcomprehension questions.

4.2.1.2.4 ZANDE TESTS

The'Zande test hadexactly the same format as theMuer test, the aajor differencesbetween the twobeingthattheZandetest had 15 nonsense syllables and the comprehensionpassage was 84 morphemes long. The disparity in length is virtually unavoidable if one isconcerned, as thetester was, with creating passages in all languagesthat have roughly thesame number of prepositions. Since Zandeuses alarge number of monosyllabicmorphemes, vhich are represented as free-standing"words" in the Zande script,an overall word count of a passagecontaining approximately an equalnumberof prepositions as found in the Bari, Lotuho and Nuerpassages would be higher for Zande.

1 5 PAGE 60

4.2.1.3 Data Collection

Four procedures were used to collectthe background, processand outcome data. Mbst of the background data obtained from Rejaf East, Bungu,Hiyala and Loronyo Primary Schoolswere collected bythe Bari (Sejario Latansio)and Lotuho (Massimino Alla.) languageofficers (employees of the Regional of Education in Ministry charge of inspecting projectschoolsand supervising project teachers) during the 1980evaluationvisits by Cziko. on the additional Bari Background information and Lotuho project classesevaluated in 1981 was also collected at this tine. Background data werecollected for the Nuer class the school in Nasir and for the at Zande classes at Yambioschool by means of the same questionnaire used for the Bari and Lotuho schools. The SIL consultants and the IRLlanguage officers completed the qmstionnaires assistance of school soliciting the officials whenever thiswas possible. Only background data from the school the in Nasir were available. However, the quality of student attendance fromthe time the Nuer schoolopened in June 1981 is extremely accurate andcomprehensive.

For the two Bari and Lotuho classes which were included forthe first time inthissecond evaluation, we attempted to collect informationon the teachers, pupils, curriculum andmaterials of each school with emphasis on particular the project literacy classes under study. This obtained via information was interviews with teachers, school directorsand school from available records. HOwever, as describedbelow, someof this information was often impossible background to obtain (e.g., pupilattendance data) or of doubtful reliability (e.g., in some cases discrepancies the teacher or director existed between what described as the language ofinstruction for a certain class and what was actually observed in the classroom). In director and/or teachers other cases, the were absent at the time of our visit. Nevertheless, the background data whichwere obtained do provide useful context of primary education information on the in the rural Southern Sudanand indicate ways in which the eight classes visiteddiffered from each other.

Process data were collectedby observing vernacular of the eight schools visited class lessons at seven (the project teacher at Rejaf Westwas on sick leave at the time of our visit). Aswas done for the 1980 impact evaluation, these data were obtained by tape recording and taking notes whileobserving a vernacular literacy lesson assistedby the language officeror SIL personnel familiar with the language who was able to providethe evaluatorswith explanations and interpretationsof the class activities.

Finally, the outcome data for the Bari and Lotuho schoolswere obtained using the literacy tests describedabove.All reading tests were administered individually to a sample of pupils from each of theeight classes evaluated. These were P3 pupils at all of theBari and Lotuho schools (exceptTorit East where P2 pupils were tested)and P1 pupilsat the Muer and Nasir and Yambio. Zande schools at All reading testsinvolved oralreading and the answering of reading comprehension oral questions. These tests were administered by therespective language officer or SIL associate whowas fluent in the local language and allresponses were recordedon audio tape. thesetests either The scoring of tookplace at the time of administration (using recordings to verify the the scoring when necessary)or at a later time from the recordings and was done by persons fluent in thelanguage being tested.The

106 PAGE 61

dictation test was group administered to all Bari and Lotuho Rejaf West. classes except The dictation textwas readto the pupils three completely times: first, through with nopausesforwriting; second, ;regressivelylarger with pauses after groups of wordsduring which time the instructed to pupilswere write what they had heard;and third, one final time pupils to check and correct for the what theyhad written. All pupils present vernacular class at the in the time of the testing weregiven the dictation addition to the test. In literacy tests, all pupilswho wereindividually tested asked threequestions: were (a) "Whydo you come toschool?"; want to learn to read (b) "Why doyou and write?"; and (c) "What do you reador write outside of school?" Pupils' answers to these tape for subsequent questions were also recordedon audio translation, classificationand analysis. At Nasir West Priaary School, the students hadbeen grouped, on the basis of past performance,into four ability levels. SIL It was the judgement of the consultantthat only thetop two groups wouldstandachanceof successfully completingthe test. For this reason, all ofthe children tested were selected from those twogroups. The children were testedindividually in a teacher's room atthe school. Each pupil was seatedbeside either the SIL consultant orthe language officer. Theevaluator, Cowan, saton side of the table, in the center the other of which was positioned the tape-recorder. The child WaS firstasked the three questions: "Why do youcome to school?"; "Why do you wantto read and write?"; and "What will youread and outside of school whenyou know how?" write Next, he was asked toread the nonsense syllables and the 20test words. Following this, he wastold toread the story aloud and after hehad finished, the examiner put thesix comprehension questions to him. The entire protocol was tape-recorded by theconsultant and the SIL consultant, whodid most of the oral interviewing, transcribedthe pupils' reading errors. Due to thefact that theSIL consultantpossesses virtually native speakercompetence in Nuer and tested becaume all of thechildren read at such a slowrate, itwaspossible to transcripts of the oral make veryaccurate reading errors during eachinterview. The Zande children testedat Yambio same procedure. Primary School were examinedusing the There were three classesformed, ability. Only the once again, on the basis of top two levelswereexamined, most of the 12) coming from children (9 of the betterclass. The interviews literacy consultant, were recorded by SIL Jocelyn Clavenger and theIRL attemptedto make a running Language Officer, Bullen, record ofthe children'soral reading duringthe test. Due to an oversight er,vrs on Cowan'spart, one irregularity occurred int the protocol: the first seven children were allowedto read the comprehension passage twice, once silently andonce aloud. This does appear to havegreatly increased the not comprehension scores of those children; indeed an inspection ofthe data indicates that the major differenceswere due to level. However, these childrendid have the benefit of tworeadings which may have improved theirperformance.

4.2.2 RESULTS

107 ' IC s " 1 4_. , 1 111 f ; I, NI 1 4 ags H l ! II !Pr , , a 11 q,igedi ilviii 0 41 d, 1, ILI,' 1 ;II;fe114 .11h141;rig m d mg 091,14111.100,1111iii di 1 ii Jay; li if ,., 1..go th1 ? hi:di 1 REIP q :ielli1111;11,1 ili .1 iq!id 1 haradiluildd101$11141111111:!I4E IIIIII Val. 0 III I I Prird.irtdiifdpi,;,roillt! Table 4,9

Background Data

School

Relit East Rejat West Bungu Variable (Barl Project) Hiyala Loronyo (Barl Project) (BarlComparison) (Lotuho Project) (Lot*Comparison)

School Data =1ftwam

Pupils enrolled In11101

PI 94 31 22 .. 39 Pl 59 30 21 .. 31 P3 48 12 3$ 29 25 P4 19 12 20

PS 35 0 25 .. 16 Pi 9 0 32 .. 16 Total 184 85 155 .. 151 Tescher1Pupll rotlo 1:28 1:21 1:26 .. 1:25 N pupils InIPS 39 ...... 42 Above puollslosing Leaving ExmoketIon 6 (151) .. . 21 (50%) First cloy of doss In 1111 5111 7115? 8115 7115 7112

(coned) 109 110 Table 4,9 (cont'd)

Background Data

School Rejaf East Rejaf West Bungu Variable Hiyala Loronyo (Bari Project) (BariProject) (Bari Comparison) (Lotuho Project)(Lotuho Comparison)

General Teacher Data

Number of teachers 10 4 6 9 6 Above teachers with

regional mother tonguea 6 (60%) 3 (75%) 5 (83%) 8 (89%) 4 (671) Education (years):

Median 10 8.5 Range OM 0 7-12 7-9 b Trained teachers (80%) 0? 3 (501) 4 (44%) 4 (V%) Experiencec

Median 8,5 6 6 8

Range 1-39 1-36 2-28 2-29

(cont'd)

112 111 Table 4.9 (cont'd) Background Data AINNII School

Rejat East kept Welt Bungu Hiyala Loronyo Variable (Bari Project) (Bari Project) (Bari Comparison) (LotuhoProject) (Lotuho Comparison)

Primary 3 Reading Teacherinformation

Mother tongue .101MIMIIMMIMMIM.,~plans Bari Bari Bari Lotuho Acholi Education (years) 11 7 12 8 Training: d TTC yes no no yes no Sileyes yes yes no yes no Experiencec(years) 32 9 1 28 3 Experience: P3 (years) 7 a 1

Primary 3 Pupil Information

En rol iment:

Total 48 12 35 29 25 Males 28(58%) 11(92%) 20(57%) 26 (90%) 25(100%) Females 20(42%) 1(8%) 15(43%) 3(10%) 0(0%) a Regional mother tongue refers to Bari in Barischools and Lothuo in b Lotuho schools. Trained teachers are defined as teachers withat least some teacher training c college attendance. Experience refers to years of working as a primary schoolteacher. d A yes in this row indicates some attendanceat a teacher training college. e A 2s. in thisrow indicates participating In SIL teacher training foruse of the project materials. fThls teahcer was the vernacular literacyteacher for project classes and the English teacher forthe comparison classes.

113 114 " Table 4.10

Primary 3 Curricula andMaterials

Teacher's Pupils' Minutts1 Language of Materials/ Subject Materials! Week° Instruction Language Language

Rejaf East School (BariProject)

Arabic 40 Arabic CME1 None Arabic

English 240 English None None

Mathematics 240 Bari None None Religion 40 Bari Catechism/ Catechism! Bari Bari

Science 40 English Missionary/ None Bari

Vernacular 360 Bari Project/ Project! Bari Bari

Total 960

(cont'd)

115 116 Table 4.10 (coned)

Primary 3 Curricula and Materials

Teacher's Pupils' Minutts1 Language of Materials! Materials! Sobject Week° Instruction Language Language

Refaf West School (Bari Project)

Arabic 180 RIM

Arts 6 Crafts 90

English 315

Geography 90 S. e

Mathematics 270 S. ileop

Religion 45

Vernacular 180 Bari Project/ =I ID Bari

Total 1170

Bungu School (Bari Comparison)

Arts 6 Crafts 40 English None None

English 400 English None None Geography 80 English None None

History 120 English None None Mathematics 240 English CMEI None Arabic Religion 80 English Bible! None English

Science 80 English None None Total 1040

117 1 (cont'd) 13 Table 4.10 (cont'd)

Prinary 3 Curricula and Materials

Teacher's Pupils' Minutts1 Language of Materials/ Materials/ Subject Week° Instruction Language Language

Illyala School (Lotuho Project)

English 225 English OUP1 None English

Mathematics 225 English RMEI None English

Religion 225 Lotuho None None

Science 90 English None None

Vernacular 210 Lotuho Project/ Project/ Lotuho Lotuho Total 1035 ...m.I.Oden.1

(cont'd)

119 120 flay tillNMI tOraPt I briollsoll

tuft% Nole Netwills1 arame 11= WIMP

holoHAW Cwation I kw Pim Nins 11104 I NON 0110 Ira Med/ WV* Nos SIMI% IN Nens

1111111111111111.11MumIMIMII

41114111 Istir IIIP seire_ 04, 1110,41 Kiwibeaus I gm. 1116110 Ibrimil IMMO WWII IP ONOwd Wank,Pro, Daft Mutt

122 PAGE 62h

Table 4.11 Background Data: Nasir West School (Nuer Project)

rrimary .1Curricula and Materials No. Pupils No. of Minutes Language of Teacher's with Pupils in Subject per Week Instruction Materials Materials Class

Arabic 240 Arabic First 0 89-110 Arabic Book

Math 240 Arabic Arithmetic 0 89-110 Book

Nuer 480 Nuer Nuer Primer 45 89-110

Religion 240 Arabic Koran 0 89-110

General Teacher Information

Number of Teachers 15

Above Teachers with Original Mother Tongue 9 (60%)

Trained Teachers 4 (2711)

123 PAGE 63

Loronyo, on sone variables, e.g., teachers with mother tongue. Although there are more teachers at Nasir West than at zany of thecomparison schools, the percentage of trained teachers isconsiderably lower--27%. The numberof subjects offered at Nasir WestIs also smaller, with the result that is spent on thee more time than at theproject schools. The SILconsultant, Marian Farquhar, Explained that she had justbegun tooffer somelimited English instruction three days a week as a response to student requests. In terms of enrollment, Nasir West appears to beslightly larger thanmany of the project schools. Interestingly, the availability of teacher'smaterials apparently exceeds that found in most project schools. The number of minutes devotedto vernacular instruction as shown in Table 4.11 represents two 60-minuteperiods per day: one before the mid-morningbreakfast andone after. Attendance fluctuated considerably from one period toanother. Many childrenwho attended the pre-breakfast class didnot return for the later period.

4.2.2.2Process Data

4.2.2.2.1 THE BARI AND LOTUHO SCHOOLS

Literacy classes were observed andrecorded forall of the Bari and Lotuho classes except for Loronyo (theteacher being absent at thetime ofour visit).Since the Primary 3 class at RejafEast was divided into two separate groups for literacY instruction, literacy classeswere observed forboth of these groups. These classes were conducted inthe vernacular for the four project classes and in English for the one comparison class observed(Bungu). As was observed in the first evaluation, the project teachers in general closely followed the instructions ofthe teacher guide for thelesson they were presenting and the same general findings discussed in thefirst impact evaluation were found again.in thissecond one. Once again, practically all classroom activities involved the teacher and the entireclass with few interactionsbetween teachers and individualpupils and virtuallyno interactions between teachers and smallgroupsof pupilsor amongpupils themselves.

Classroom activities in all classes observedconsistedprimarily of group reading and repetition of syllables, words andsentences (oftenunrelated sentences). At the four project schools there . was also some practice in spelling and writingindividual words-and inthe making up of new sentences containing words included in the lesson. The only observedactivity which clearly involved readingcomprehensionwas theanswering of reading comprehension questions based on .the lesson'sstory whicheach pupil read individually and Silently. The entire classat thecomparison school consisted of repeating four English sentences writtenon theboard. These sentences were:"A man and a womana boy and a girlare walking. They are walking to a house. A tree is near the house. The man and thewoman are walking around the house." These sentenceswere either read aloudby the entire class.with the teacher pointingto eachword as they were read or read aloud by individual pupils who came up to the chalkboardand pointed to each word as they read. These activities lastedfor about 20 minutes until itwas decided to use the remaining classtime toadminister the dictation test.

12 4 PAGE 64

There wasno evidencethat the pupilsunderstood the meaningofthese sentences, although the teacher did review themeaning of the wordhouse at the beginning of the lesson.

As notedthe previous year, a great dealof tinewas spent inboth the project classes and comparisonclass by the teacher in writing syllables, words and sentences on the chalkboard. In the case ofthe four project classes, what was written on the chalkboard -was always already printed in the pupils' books.

In one of the project classes observed, the teacher never called the pupils by name and it appeared that he did not in fact .know their names. In the other classes observed, teachers also often simply pointed to pupils to select them for individual reading activities, althoughthey did on occasion use the pupils' names.

It vas alsoobserved that in generalthe project pupils appeared to read more fluently. In theliteracy classthan they did duringthe individually administered oral reading tests. It is notknown whether this wasdue to being familiar with the materials of the class lesson, anxietycaused by the testing situation, or a combination of these or other factors.

4.2.2.2.2 THE NUER SCHOOL

Three different lessons were observed atthe Nasir West School. As stated earlier, there:were about96 pupils whoregularly attended theliteracy classes. The SIL consultant, Marian Farquhar,devised a unique solution to deal withthese large .classes. Sherecruited two untrained teacher helpers from the school. One was a woman who had been trainedby the SIL consultant before the civil war and who was literate in Nuer. The other, aman, was a messenger, who-had learned to read Nuer in the church literacy class.The SIL consultantdivided the'entireclassinto four ability groups.. The IRL literacyofficer, Simon KuoonPuoc, taught the highest level, the SIL consultant (who, as pointed outearlier, is bilingual in Nuerand English) taught thenext highest level and the teacher helpers taught thetwo lower levels. Theteacher helpers learned some procedures simply bywatching the IRL officer teach somelessons, such as the letterlessons, toa combined class made up of all four groups and they also were periodically supervised by the SIL. consultant. Pupils could moveup to ahigher group or be asked to join a lower group wheneverthe IRL languageofficer or theSIL consultant felt this "was appropriate, and, indeed, the Cowan Observed thishappening several times during his stay.

The procedureused for presenting each new alphabeticalsymbol ("letter" lessons) followed a given sequence, which the teacher workedthrough in five steps which were intended to present the new letter and contrast it with known sequences. This was a laborious process which requires writing the letter and contrasts. (represented in five boxes in the primers) on the board and having the students read.thei aloud in chorus and, to a limited extent, individually. Only steps one and five aCteally contained meaningfulmaterial. In general, one cannot avoid the impressions thatthis activity, which is the cornerstone of the method, was time-consuming and boring. Only a small percentage of the students (the betterones, who may alreadybe able to read the letter) were capable of maintaining their concentration for the 30 to 50minutes that may

125 PAGE 65

Attention wandered during thetime that elapsed be expended onthis lesson. on the board(the chalkboard at this while the teacher wrote each new box of the letters were notdiscerniblefrom the school was so slick that many teacher and during the endlessChoral repetition after the back of the room) The teacher did not read whatever sequence is on the board. or a student has so aftertwo orthree often deviate from oneorderof presentation, to memorize each boxand repeat an entire repetitions, it was possible sequence with one's eyesclosed. to all four ability groups, At Nasir West, the letter lesson was taught worked on whatever lesson which then broke up intotheir individual groups and While Cowan wasthere, he watchedthe entirestudent they had reached. finished, the population being' taught letter lesson 42.As soon as this was Only the most advanced group students reassembled in theirindividual groups. the lesson they had continued working on lesson42. Group 2 dropped back to other two groups worked at evenlower been working on lesson19and the followed at Yambio School. There the three levels. (This pattern is not joint meeting takes place. This say groups work atdifferent speeds and no obtained atYamblo to be have something to dowiththe superiorresults discussed later.) at Nasir West The reading lessonobserved using the comprehensive passage advanced pupils - -Group 4. There were 12 children in School was with'the cost out tothe The procedure was asfollows: Bookswere passed this group. subgroups of four The teacherhad the children dividedinto saall children. aloud in chorus the five'boxesin the students each. Each subgroup first read When each subgroup hadfinished, the teacher toll the letter drillpage. their books to silently. They did this, closing class to read the story the finished. When, all but one student had finished, indicate when they were questions about teacher informed the classthat.he wasgoing to ask them some boy, whoanswered it In the firstquestion, he called on one the passage. question The entire class blurted out theanswer to the next correctly. it. When all of the before the child called onhad an opportunity to answer into two groups and questions had been answered, the teacher split the class aloud in chorus, sentenceby sentence, alternating the entire story was read then repeated, but this the other. The choral reading was from one group to the remaining time, one group read halfof the passage andthe other group where the first group had the There .followed another choralreading, half. previously and vice versa. The half that had been assignedto the other group devastatingly monotonous rhythmic chanting quality ofthis choral reading was that it did not appear toapproximatethe true and unnatural,. in the sense readers that pauses characteristicof thefew Nuer patterns of stress and rhythm in Again it appeared that thesounds of the words and the Cowan heard. importance thanthe they were repeated sees to be given much more which of literacy. When the meaning of the words and the communicative function individual children to choral reading was finished, the teacher called on hour and ten minutes. read. The entire lesson consumed one readingworkshop held in The only innovation developed asa result of the 1981 and incorporated in theCycle2 materials at that Neridi in November of cloze exercise, the drill. This was essentially a time vas the "context clue" different levels of contextual purpose of which wasto show children that decipher words they may notyet have infermation are available in thetext to above, was that thecontext encountered. The original plan, as described PAGE 66 clues for lessons 1-25would focus items tofill the on syntax; for exercise would lessons 25-50, lesson 51on, the be semantically the choiceof alternatives for determined and (visually)similar. the clozeiteas would that from at the Initial reports be graphemically testing sites from the SIL drills in Visirand Yamblo consultants andIRL officers were extresely indicated that the IRL language popular withthe these contextclue officer provided children. At the Nasir taught.For a demonstration Vest School, the purposesof this of how instead of demonstration, these unitsare teaching thelesson froa one adjustment was taught from the books,as is normally was made: the chalkboard done, the to all fourlevels. lesson The teacherbegan by and replaced writing theframe by a blank)and the (a sentencewith one He thenturned to the three possible word deleted one of these children andsaid: choices (words)on the board. three words- "I wantsomeone to read blank space." -the onethatis the this and put Almost allof the right (correct) hands. There was entire ane - -in the visible group (48children) "qu p, qu excitement inthe room, raioed their: r, qu r!" ("teacher, many children fingeys inan atteapt calling out to get the teacher, teacher!")and snapping teacher tochoose thee. their The teacher, Simon Kuoon well likedby the Puoc, wasextremely students. He skillful andwas obviously the variousoutcomes of developedan effective the sentence the students' technique forhandling and simply attempts. Frequently three substitutea word in the a child wouldread alternativesarrayed below blank thatwas not wouldrespond: "Okay. in thesentence. one of the Isthat word In thiscase, the substituted,e.g., (says the word teacher "cow") oneof these that the.child has (pointing .tothe first three wordshere? word). . Is it respond: "No." The class,or at least this one?" "Is it thisone?" (same part of response). "No, it response). it, would isn't anyone of these "Is thisone 'cow'?" call on.another student. here." Then (same As soonas z sentence he wouldproceed to alternative, theteacher would was read with was correct, take painsto first the correct usually saying ask the classif the right?"Upon receiving something like,"Did Col (the answer each alternative an affirmative child's woe)get it and demonstrate answer, he would as a possible that the go back andtake up context ofthe sentence the children, substitution. Initially, the excluded it 'inserting teacher read then asking: thee intothe blank, the alternativesto ""Is thisNuer? Can reading theentire sentence response formthe class we say thisin Nuer?" and Cowan's would belaughter and Thepredictable suggestion, the teacher a chorus of"No." volunteer toread the changed theprocedure and Later, at Into the incorrect hadthe children blank andask alternatives;then he would the children substitute This latterprocedure if theresult were them. the lesson seems to be acceptablein Muer. somewhat, it more desirable,since even nonethelesssolicits wider though itextends 4.2.2.2.3 participation. THE ZANDESCHOOL The Zande school atYambio had materials interms of an edgeover the testing at Nuer availableresources. site for theNuer Vest School,whereas There wasonly one been made to there werethree at trained teacher group the students Yambio. The decision to the twohighest by abilityand to assign had groups and theother to the the two best consultant,Alice Van most eleaentary teadhers Bergen, then class. consistentlywith their monitored all The SIL teachers atthe pace classes and determined by worked the teachersat the

127 PAGE 67

time of the consultant'svisit. The lowest groupwas still on the first book, the middle groupon Book 2, lesson 29 and the top group was readinglesson 49 in Book 3. lbe consultantobserved three different which were taught in kinds of lessons atYambio the two advancedclasses.

Perhaps the biggestdifference between the procedure for teaching the letter drill witnessed inlasir and the one used in emphasis.placed on syllabificetion. Yamblo wasthe extra The teachers in Yambio the letter lesson in tended to complete about 35 minutes, whichis probably the maximum time that can be devotedto this lesson without amount of was still .too muCh choral asking it deadly dull. There repetition and virtuallyno meaningful activities associated withthe presentation of the letterlessons, butthe consultant witnessed some effectiveword attack teaching students had in the functor drills. When the trouble reading a bisyllabic word, the teacher would word on the board,,cover the second syllable write the the first syllable. with his hand and havethen read Then he wouldreverse the procedure and read the second syllable, have the children whereupon he wouldthen present the the children wouldbe ableto read entire word and it. This procedurewas extended effectively to trisyllabic'words beginning at syllables and amalganating the left and uncoveringthe new them in successionfrom left to right, bakere, *great", the teacher e.g., for begins, "ba"; the children repeat teadher reads "ker"; the this, then the Children repeat; theteacher reads "ba ker", children repeat. The pupils are then and the asked to read theentire word. The other major difference -seen at Yamblo participation. was more individual Althoughthe individualclasses are about both testing sites, the samesize at there seems to bemore provision for individual the Zandeclasses. This is no doubt reading in due tothe factthat there are trained teachers at Yambioand the Zande sore children have better overallreading proficiency; hence, individualparticipation does not it does at Wasir. take up as much timeas

4.2.2.3 Outcome Data

4.2.2.3.1 BARI AND LOTUHO LITERACYTESTS

The results of the' Bariand Lotuho literacy tests arepresented in Tables 4.12 and 4.13.The effect sizes (ES) were calculated by subtractingthe mean of the comparisongroup. from the mean of the project group anddividing this difference by-the standarddeviation of the comparison effect size group. Apositive indicates that theprojectclass was superior to class while a negative the comparison effect size indicatesthat the comparison superior. An effect size with class was an absolute value of.5 or greater can be considered indicative ofan important difference (Minium, 1978,p. 96). For theBari schools, examination of Table 4.12 revealsthatRejaf East scored above thecomparison school (Bungu) on only three of theeight literacy measures while Rejaf Westperformed worse than eight measures. comparison classon all of the

128 Table 4.12 PAGE 67a Bari Literacy Test Results

Schoola Rejaf East Testb'c Rejaf West Bungu (Project) (Project) (Comparison) 1. Vowel Words (10) M 3b 8.2 7.4 9.1 2.7 4.0 1.7 TT -.52 -1.00 Tr 20 2. tonsonant Words (19) 11 15 M 14.5 3b 14.4 16.1 4.2 5.8 1.2 TT -1.33 -1.42 ir 20 3. 'Known Words (10) 11 15 M 9.4 3b 8.4 9.0 1.1 2.0 1.2 'ET .33 -.50 -- ir 20 4. Dnknown Words (10) 11 15 M 3b 6.6 5.9 7.2 3.3 3.6 2.6 TT -.23 n -.50 20 11 15 5. Total Word Skills(49) M 3b 38.8 36.1 41.4 10.3 14.7 5.1 IT -.51 -1.04 n 20 6. 'Story Words (52) 11 15 M 44.2 TD 42.2 45.0 TY 14.2 14.5 10.4 77 -.08 -.27 20 11 7. !leading Comprehension(6) 15 M 3b 3.2 2.4 2.7 1.7 1.2 1.4 TV .36 -.21 -- ri- 21 8. Dictation (16) 11 15 M 15.0 TD - 13.7 1.4 -- 1.5 'ET .87 -- n 24 -- 28 aPrimary 3 pupils were tested at Hiyala and Loronyo;Primary 2 pupils were bNumberstested at Torit East. In parentheses indicate variable. the maximum possiblesocre for each test cM = arithmetic mean, SD= standard deviation, n S) was computed briubtracting = group size; effect size the project the mean ofIhe comparisongroup from group mean and dividing thisdifference by the standard deviation of the comparisongroup.

129 PAGE 67b Table 4.13 Lotuho Literacy Test Results

Schoola Hiyala Testb'c Torit East Loronyo (Project) (Project) (Comparison) 1. Vowel Words (5) M rD 4.2 4.9 2.9 1.7 .3 2.7 Ts. .48 n .74 20 10 2. 'Consonant Words(5) 7 M 3b 19.8 18.6 9.6 7.7 4.5 9.2 IT 1.11 .98 n 20 10 3. Known Words (10) 7 M 9.0 9.1 SID 3.3 2.2 2.2 4.3 TT 1.32 n 1.35 -- 20 10 4. 'Unknown Words(10) 7 M 3b 7.3 6.3 2.9 3.3 2.9 4.9 Tr .90 .69 W.' 20 - 5. Total Word Skills (52) 10 7 M llb 40.4 38.9 18.6 14.0 9.2 19.1 Tr 1.14 1.06 n 20 - 6. Story Words (53) 10 7 M 43.2 43.1 31) 15.7 16.7 11.5 25.5 TS" 1.08 n 1.07 20 10 7. Reading Comprehension(8) 7 M 6.0 3b 4.7 1.4 2.6 1.9 2.5 'ET 1.84 1.32 IT- 19 8.Dictation (20) 10 7 M 14.9 7.2 TD 3.2 6.6 7.0 5.9 ES 1.98 IT" .68 -- 22 42 17 aPrimary 3 pupils were tested at Hiyala and Loronyo;Primary 2 pupilswere bNumberstested at Torit East. in parentheses indicatethe maximum possiblesocre for each test cMvariable. = arithmetic mean, SD= standard deviation, n DS) was computed = group size; effect size by-Subtracting themean oflhe comparisongroup from the project groupmean and dividing this difference deviation of the comparisongroup. by the standard PAGE 68

Among the Lotuho pupils (seeTable 4.13) the two Lotuho outperforned the comparison project classes class (Loronyo) on alleight measures of literacy skills with all effectsizes of .5 or above. Of thetwo Lotuhoproject classes, Blyala performedbetter than ToritEast on while Torit East scored six ofthese measures slightly higher than Hiyalion measures. the remainingtwo

It should be keptin mind that the pupils tested at Torit Eastwere in P2 while the other two classestested werein P3. of Torit East have acquired This suggests that thepupils literacy skills ata much faster rate than the students in the other"twoschools. Although we can only is the case, the speculate on why this fact that the town ofTOrit is much more developed rural villages of Hiyala than the and Loronyo isvery likely related to success of these pupils. the relative In contrast tothe villages of &Valeand Loronyo where children arevery unlikely to see anything Torit (which became written outsideof school, the provincial capitalof Eastern Equatoria at least some signs with in 1981) has written English and/or-Arabicand one can also find canned foods and otherpackaged goods in the labels. market and shops withprinted It is also fairly certainthat some of from higher socioeconomic the Torit Eastpupils come baOkgrounds than pdpils ofthe other Finally, it should be two schools. mentioned that the P3vernacular teacher at Torit complained that on the day East his pupils were tested,many of his best students were not present in^class sincethey were making the head of the regional preparations for the visitof government which took placethe following day.

Both the Bari ahd Lotuhoteat results are results obtained the fairly consistent with the previous year. In boththe first and secohd evaluations impact the Lotuhoproject schools clearlydid literacy tests than did the better on the Lotuho comparison school whileamong the Bari schools the project schools did notgenerally do better than the comparison school. These findings were obtained inspite of thefact that although for the second evaluation e, sought to retest all childrenwho had been individually year before tested the inconsistent attendance madethis impossible. who were telited individually (Of the 14 pupils in Rejaf East in 1980,only 5 or 36% were individuallytested in again 1981 with the addition of 15 pupils individually who were not tested in 1981. For Bungu, Hiyala end Loronyo percentage of P2 pupils schoolsthe individually tested in 1980who were again tested individually in 1981 at P3was 83%, 22% and 44%, respectively.)

Therefore, if,we assume (as appears to bethe case) that mostprinary schools in Southern Sudanhave no or very limited for use amounts of pupil materials in teaching vernacularliteracy skills, then the the Lotuho project, and ,difference between Lotuho comparisongroups in perforianceon the Lotuho literacy tests could betaken as some indication of the impact of theliteracy project. However, as notedlast year and seen again in findings, the thisreport's relatively goodperformance of theBari suggeststhat what makes comparisongroup a difference isthe availability and vernacular literacy materials useof by the pupils and there is as evidence that the Cycle yet noclear 1 materials are superiorto the had been used by the Bari older materials that comparison group (Bungu) at P1and P2. In fact, the generally superior performance ofthe Bari comparisongroup in the total absence ofvernacular literacy instruction during P3 indicatesthat the naterials and methodused for this groupin P1 and P2 may havesome

131 PAGE 69 4; isportant advantagesoverthe newproject Cycle 1 described in materialsand methods. As the first evaluation,one of the major differences between the two materials andmethodsis the teething of letter to decode written names and sound blending words to theirspoken fOrms. This would drills and format ofthe new Bari and suggest that the Lotuho materials designedto decoding or "word attack"are not as teach a effective as the moretraditional techniques of teachingletter names and the sounds represented byeach letter. The overall literacyskills of the Bari to remain quite and Lotuho projectgroups appeared low. Very few of the Barior demonstrated literacy skills Lotuho pupilstested that went beyondbasic word recognitiOn attack skills.Although it is Cziko's and word impressionthat there was ingeneral a slight improvementin these skillsover last year, pupilscontinued to have considerable difficultyreading most words which they had not alreadystudied, including common shortwords sade up of letteri have been well familiar: with which the childrenshould Since only a few Lotuhopupils demonstrated fluent oral readingskills with good fairly cosprehension, it appears thatthe majority of the projectpupilstested have acquired atbeat somebasic decoding and word recognitionskills but are weak skills. It may well be that in reading comprehension the hdgh repetitionrates of a relatively number of trords and the slow rate of small introducing new wordsintothe Bari materials noted earlierare at least partly children are not given responsible for thissince the such opportunity todevelop skills for words.Considering the amount decoding new and type of vernacularsaterials these pupils have available .to thes,the type of instruction their 'very they receive in schooland lisited or non-existent exposure demonstrations of the outside of school to functions of literacy, we would present conditions the predict thatunder .overwhelling_majority ofthe pupils inthese Bari and Lotuho classes,will notachieve true fUnctional do .will not keep literacy and that thefew who these skills forlonggiven the almost vernacular reading materials total lack of inthe ruralareas of the Recommendations to improve SouthernSudan. this situationare included in Chapter 5. 4.2.2.3.2 MUER LITERACYTESTS

Table 4.14 shows that the perfOrmance ofthe Nuer children tests was not very impressive. on the literacy A generally decreasingtrend is evident in the percentages of thenonsense syllables (72%), known words (67%),unknown words (25%) and number ofcomprehension questions (27%) correctlyanswered. One poSsibleinterpretation of this isthat, atthe' time the administered, thechildren had gained test was symbols into sound (evidenced some minimal facility inrecoding by thechildren's performanceon syllables), butthat they ware still the nonsense unable to applythis to new words hence werelargely dependent and upon recognizing known wordsin order to read with comprehension.Two findings tend to the support this interpretation. First, Pearson product-momentcorrelation between the correctly read in the number of knownwords tests and the numberof comprehension correctly answered isquite high, r questions .72. Secondly, theperformance of the children in the 'oralreading test (see Table 4.15)mirrored the classic symptoms of poor readersfound in studies like Pugh (1977). Weber's (1970)and Neville an

Whenever the children encountereda wordthat they did not recognize, they would adoptone of two strategies: immediately most of the time they

132 PAGE 69a Table 4.14 Nuer Literacy Test Data:Nasir West Primary School

A. Number of Pupils Tested 21 B. Average Age 9 yrs.8mos. C. Average Number of DaysAttended Class (total =115days) 87 Range 78 - 112 D. Mean Number of NonsenseSyllables Correctly Read(12) 8.7 SD 3.0 Range 1 - 12 E. Mean Number of Know Words Correctly Read(10) 6.7 SD 2.6 Range 2 - 10 F. Mean Number of 'Built' Words Correctly Read(10) 2.5 SD 2.8 Range 0 8.5 G. Mean Number of Words in Passag Read Correctly(56) 33.11 SD 12.3 Range 11 - 50 H. Mean Number of Comprehension Questions Correct(6) 1.6 SD 1.3 Range 0 -

Note.Numbers in parenthesesindicate the maximum possible each test. score for

133 PAGE 69b

Table4.15 Nuer Oral Reading Errors

Error Type Examples of Total Number of Total Oral Reading Errors

1. Internal Vowel Change 34 9.3 ken kin reit rok 2. Substitution 53 14.5 tee -0 gat r 4 bithdek

3. Final Element(s) Retained 65 17.8

EIDO4k:11 Dr) pi 4 ki kaa r 4 raar

4. Initial Consonant Retained 86 23.5 bake! * ba guath 4 gua bok 4 born

5. No Attempt 45 12.3 6. Single Vowel Retained 52 14.2 rek c Viy6 4 n6Q kula r 4Cal,

7. Difference of one Vor C 29 (Includes deletions and 7.9 epenthesis) ken 4 kent kuaa r 4 kfil r wecdicn 4 wecie n

8. Other 2 .6 ji egella

gap 1 se a

I I I ii.laj 1 1 ; I . 4 4. i np sif iii si 1 . z 1116111 ill: !MIMI! I PAGE 71

did a child ver backup and attack a sentence again in an attemptto make use ot previous contextto deciphera problem word. Taken together, these two strategies arguethat theNuer children areextremely "wordoriented," and have yetto develop proficiencyin word attack skills. Moreover, they sees unable to make *Motiveuse of context in reading. Pearson product-moment correlations were computedto determine possible relations between performanoeon different levels of Item were all fairly the test and attendance. low: (a) attendance and (b) attendance and comprehension score, r = .51, nonsense syllables, r= .40 and (c) attendance words, t .11. and known

4.2.2.3.3 WIDE LITERACYTESTS

Due to the projected short duration of thesecond at Eambt0, testing impact evaluation visit on a scale comparable tothat undertaken at oontemplated. It was hoped Nasir was not that there would betime to adequate Zande reading develop a reasonably test similar to theone used for the Children for the first Bari and Lotuho evaluation which would be ready relieving year's third for use forthe iapact evaluation.Other than that, Cowan had to observe severalclasses and collect planned processdata. However, these plans were changed when theperformance of the first classobserved revealeda dramatic contrast withwhat the consultant had seen in Nasir.The unavoidable ingression was that theZande pupils were reading with fargreater facility than the best Muerpupils in Nasir. The contrast wasso strikingthat it cried out for anexplanation, particularly since bothtest populationswere, to a large extent,matched, i.e., they had both begun school atthe same time, had been assigned toclasses according to their ability andwere roughly at the same level in theirprimers. There was, one The lands children had very significant difference. been allowed totake their primershome after school about five weeks prior tothe second impact evaluation visit, whereasthe Nuer children had access totheir piimers only for thetwo periods theywere in school. The possibilitythat the Zande children's superiorperformance was due to extraexposure to reading outside of class was an obvioushypothesis. In orderto deteraine whetherthis impression oftheZandepupils' proficiency was justifiedand to discover whether reading outside of class, the children were indeed Cowan allocated differenttesting tasksto the IRL and the six personnelpresent. The IRL language officer,Bullen and the SIL oonsultant, Jocelyn Clavinger,adainistered the Zande reading earlier to individual test described students while theconsultant; IRL Officer Assn and SIL consultant John Baptist Alice Van Bergen performedspot tests with pupils each of thetop two classes. in The procedureforthe latter follows:we askmd the entire class three test 1183 83 questions: (a) How many of your books at home?(b) How aany of you read you receive help withyour reading at home from your parents? and (c) How many of can reed are sand by them you who don't have parentswho todemonstrate how wellyou read and write regularly?Inthe tap class (21 children), 11 children claimedthat they received help regularlyfromtheir parents, six read by themselves children claimed that they regularly and 20 childrenclaimed required them to demonstrates that theirparents what they had learnedin school regularly. child said that she hadnot read ahead One because she doesn'thave a primer. In the second group, all 14children present that day claimed to dosome reading

136 PAGE 72

in their books at home. Eight out of14 children claimed to end of the primer. have read to the

Next, we asked individual children how far beyondthe primer lesson taught that daythat they had read. As each child responded lesson, he or she with the numberof a vas asked toturn to thecomprehension passage lessonto read it aloud and for that then to answer thecomprehension questions followedit (theSe were asked that orally by the teacher). The oral readingerrors were recorded.as,wellasthenumber of comprehension answered.- Of questionscorrectly the.19.children.in the topgroup, 17 were able to read at one lessonj:beyond the Currently least completedlesson (lesson 49). Two children attempted lessons whichwere too difficult for, them the story completely. and were unable to read The breakdown of the lessons attemptedby the students and number.of 'comprehensionquestions answered successfully 4.16. Note is shown in Table that.ever 25% ofthe students werecapable of reading.12 lessons beyond.where. to 15 the class actuallywas in the primer and 77% of wascapableof 'reading at the class least threelessona beyond the completed. In the next lessonjust lowest class, Group 2, 50%ofthe students were able to read four.lessona_beyondwhat was currently being to read three lessons taught and 60% wereable beyond the current lesson.The mem oral this class was higherthan error rate for the top group (Group1, M = 3; Group 2, and the children M = 5.2) in this class tended to block (refae to couldn't read, a word) persist Mien they more frequently than theChildren in Group One is inevitably. tempted .to combine the resultsof thesespot tests with the results of thereading tests administered by JocelynClavinger and Bullen (see Table 4.17)- tothe 12 ZandeChildren and make some Muer comparison with the children tested 'inNasir. Strictly speaking, specious, since there.is such a comparisonis no:way of determining whetherthe tests devised for the Nuer :and Zandechildren are in any way equivalent. small salple ofchildren Nevertheless, the tested by Jocelyn Clavingerand Bullen achieved much higher scores 'on.all parti,ofthe test than test. the Wier children didon their Table 4.18 shows thatbasically the sametypes of errors sone exceptions. As one might:expect occur with in a language which containsa lot of initial consonantClusters comprised of velars and bilabial stopsor nasals and a homorganic atop,like Zande, an epenthetic vowel /a/ will be insertedin reading, e.g.,error type 2. Similarly, initial segments error type 1. will be deleted:

It is difficult not toconclude that the Zande Children were readingbetter than their Muercounterparts. :ifthis conclusionis in fact true, difference between the two the groups could beattributable toone or following factors. more of the Firstand most obviously, theZande Children have literate environment. a more Because they are allowedto take their books home, possibility for additional the reading outside of classexists (and is place, if. we are to :believe taking the .children'sreports), whereas the Children's reading experience Nuer is confined to two periodsat Second, the school each day. Zande schOol has moretrained teachers. Children can receive Consequently, the more individual instructionin their ability pointed eut earlier, in Nasir, groups. As all of thechildren are forsome period of the day taught by one teacher. It is very possible that taking place in these classes less actual learning is than inthe smaller groups, where be mOre focused. attention may Athird possibility is thatthe tests are not of difficulty.More time was spent equal developingthe Nuer than theZande test.

137 PAGE 72a Table 4.16 Zande "Spot" Reading Test

Student No. of Questions Sex Lesson A ttempteda Correctly Answered

Student 1 male 50 All Student 2 female 50 3 out of 4 Student 3 male 50 Ali Student 4 male 51 All b Student 5 female 52 3 out of 5 Student 6 male 52 All Student 7 female 53 All b Student 8 female 54 3 out of 4 Student 9 female SS All b Student 10 male 56 All (considerable difficulty) Student 11 male 56 All Student 12 male 57 All Student 13 female 59 Failed to completepassage Student 14 female 59 Failed to completepassage b Student 15 male 61 All Student 16 male 62 4 out of 5 b Student 17 male 62 3 out of 5 Student 18 male 63 All Student 19 male 64 All aThereare 64 lessons in the primer. b These children had some trouble with the story.Their reading was slower and more halting thanthat of the other children the average number oferrors (3). and they made more than

138 PAGE 72b

Table4.17 Zande Literacy Test Data:Yambico Primary School

A. Number of Pupils Tested 12 B. Mean Number of Nonsense Syllables Correctly Read (15) 13.8 SD 1.8 Range 10 - 1 5 C. Mean Number of Known WordsCorrectly Read(10) 9.8 SD .4 Range 9 - 10 D. Mean Number of "Built" WordsCorrectly Read(10) 8.2 SD 2.7 Range 1 - 10 E. Mean Number of Words in PassageRead Correctly(84) 81.2 SD 2.7 Range 72 - 84 F. Mean Number of ComprehensionQuestions Correct(7) 5.3 SD 1.8 Range 3 - 7

1 39 PAGE 72c

Table 4.18 Zande Oral Reading Errors

Error Type Examples 11, of Total Number of Total Oral Reading Errors

1.Deletion of Initial Consonant(s) 7 12.3 gbe 4 be ngbarago gbarago barago

2.Epenthetic Vowel 11 19.3 mbi mabi gbe gabe 3.Final Element(s) Retained 7 12.3 ngbarago rungo bere 4 mere 4. Initial Consonant Retained 12 21.0 oro 4 ora bayugupai bapa 5. Deletion of Medial Consonant 3 5.3 gunde gude bambiko bakiko 6. Epenthentic Consonant 5 8.8 gbe ngbe 7.Substitution 3 5.3 8. Retention of Single Cor V 2 3.5 nda du 9. Internal Vowel Change 1 1.8 bere bire 10.Nd Attempt 6 10.5

140 PAGE 73

Fourth, the inconsistentprocedure used in administering first seven children the Zande test to the may have elevated theirscores somewhat. always the possibility that Fifth, there is the difference inthe samplesizes ofthe two groups testedmay havecontributed to the difference in performance. Had there been more time, the testing of more Zande children mighthave brought down the overall performancelevel ofthat group. with which words are And finally, the frequency recycled in the Zande primersay be greater than in the Muer priser, making the formereasier to read.

It would appear thata combination of the first two the Zande children becoming factors has resulted in better readers than the Nuer children. this conclusion deserves Obviously further investigation.Still, itappears important to point out the importantimplications which arise from Zande and Nuer children. the comparison of the Be have argued earlierthat a crucial difference the setting of this in project and most otherliteracy endeavors is rural Sudanese children have that the little or no exposure tothe written word outside of school The difference in reading betweenthe Zande andNuer children suggests that the single mostimportant factor in promoting increased exposure literacy say be an to reading . materials outside of school and the encouragement and assistance they report receiving from implies that the their parents.This effectiveness ofthe .current programcould be dramatically by the institution increased of three policies: (a) allowing the to take their books home, children (b) encourgaging thento read outside of schooland (c) creating 'amoreliterateclimate inthecommunity through the dissemination of books andpamphlets in the local languages whichcontain subject matter of interest toadults and Children.

4.2.2.3.4 PUPIL INTERVIEWS

Answers tothe threequestion asked of those students individually tested who were were categorized for analysis. In answering question, "Why do you the first come to school?", the majority of the pupils in all the Bari and Lotuho schools except Hiyalagave"to learn"as the reason, with reading or writing principal most commonly claimedas the particular skills they wished tolearn. Among the Hiyala pupils, pupils interviewed however, almost half of the gave "becoming civilized,wise, and/or clever" reasons for attending school. as their Only one ofthe Bari and Lotuho pupilsgave an employment-related reason for attending school anda total of reported that they five students felt obligated togo to school. Both theNuer and Zande childrenseemed tobe somewhat confused by the firstquestion. The Nuer children gave differentanswers, ranging from "my brother want to know all things broughtme," to "I they teach in school, becausesome day I will be a big chief." The majority, 8of the 18 Nuer children who respondedand 10 of the 12 Zande children simplysaid "to learn."

For the second question, "Why do you want tolearn to readand write?", most pupils reportedthat they wanted to become literate so that they could becomeimportant(often referred to as being a"big man"), civilized, knowledgeable,,clever, or "likeyou" (referring to statusof Cziko and the relativewealth and the Lotuholanguageofficer, Nassimino conducted the interviews). Allan, vho A fair number of. employment -related reasons were also given at the three Lotuho schools. The Lotuho pupils relatively more education- alsogave and wealth-relatedreasons for learning to read and writethan did theBaripupils. The Nuerchildren gave widelyvarying

141 PAGE 74

answers. Five children said theywanted to be able to read the B!ble. The remainder gave answers rangingfrom "So that when Isee something that is written, Ican read it, to"So that I can have respect for myclan."The Zande children, on theother hand, were split evenly in their responses. Six children said simply "to know,"but five said "to be a person of thefuture." This may be interpretedas meaning to be soseone who is thus has a better opportunity knowledgeable and who to improve hislife. One child said "I be ableto read so want to that if myfather dies in a far place and writes letter to tell se about it me a I'll be able to understandit.

Finally, when asked, "Doyou read or write outside of school?"answers varied widely from schoolto school. All the pupils interviewed Torit Eastreplied that at Hiyala and they did read or writeoutside of school, while one-fifth to one-third of only the pupils in the remainingschools reported doing so, except for Rejaf East, wherehalf of the pupils reador write outside of school. interviewed said they did Among the materialsreported asbeing available at home to read,school books (the new used at vernacular reading materials the project schools) appearedtobe the most popular, religious materials followed by in the.vernacular languageor in English. put tothe Nuer and Zande children The question differed slightly dueto the fact that these pupils were all in' PI. The actual phrasing of williou read or write when the question was: "What you learn how to?"Again, the Nuer children quite varied responses.' Four gave children said that theywould read "some books." No "said that they Wouldread the Bible. to be able Three children said thatthey wanted to write a letter or whatever they felt like. would get a job with his One said that he knowledge and writingwould help him in five children gave this; and responses Which were moreappropriatefor the question: two said they would become second pastors and teach theirpeople and three said that they would becometeachers. Elevenof the Zande children saidthat they would read the booksthey had in school and would write the Zande language. one simply maintained that he

4.2.2.3.5BARI ORAL READING ERRORANALYSIS

In order to obtain a better understanding ofthe readingprocesses and strategies used by the children, a detailedanalysis was undertakenof the oral readingerrors made by the Bari project and comparison pupils. First, the oral reading performanceof each pupil each was transcribed bylistening to tape-recorded performance andnoting all deviations story on from theprinted copies of the text. These errors included all order changes, substitutions, word repetitions, deletions andinsertions. transcript was The accuracy of each then verifiedby the scorer bylistening again recordings and making changes to all the where necessary. The scorer then filledout an error classification sheet foreach pupilwhich described each error and its frequency according to oral reading the following classificationscheme. VORDSUB: Substitution of the printed word by another Bariword. NSSUB: Substitution of the printedword by an Bari, English, or Arabic utterance which isnot a word (NSSUB = nonsensesubstiturpn). ORDER: Change in the order of twoor more words of the text.

REP: Repetition of a wordor phrase of the text.

142 PAGE 75

DELETE: Deletion of a wordor group of words of the text. INSERT: Insertion of extraneouswords into the text.

SIMSPELL:WORDSUB or NONSUBbeginning with the word it replaced same letter as the text or the rest of spellingjudged very doto). similar (e.g., toto for

NC: Noncontext error, i.e., any error other than NSSUB or REP which not conform tothe preceding did syn- tactic and semanticconstraints of sentence in which itwas located. the

PC: Preceding-context error, i.e., any error other than NSSUBor REP which conformed to the precedingsyntactic and in which it was semantic constraints of thesentence located but did not conformto the sentence. constraints of the entire

SC:Sentence-context error, i.e., any errorother thanNSSUB or REP which conformed V:, the syntacticand semantic constraints of entire sentence in which it wez locatedbut did not conform to thediscourseconstraints of all preceding sentences of thetest.

ix: Discourse-context error, i.e., any error other thanNSSUB or REP which conformed to thesyntactic and semantic it constraints of thesentence in which was located as wellas tothe discourse constraints aentences. of allpreceding

NSSCOR: Correctednonsense substitution, i.e., aNSSUB which was spontaneously corrected bythe pupil.

MCCOR: Corrected NC error.

PCCOR: Corrected PC error.

SCCOR:Correcied SCerror.

DCCOR:Corrected DC error.

NS: No-source error, i.e., any NC error whichwasnot SIMSPELL and consequently could not beattributed to a pupil'suse visual information. of either contextualor

MS: Multiple-source error,i.e., any PC, SC, or SIMSPELL and consequently DC error whichwas also could be attributed to theuse of contextual and/or visual information by thepupil.

TOTAL:Total nuaber oferrors, i.e., WORDSUB + INSERT. + NONSUB ORDER + REP+ DELETE

TOTNONREP: Total number of nonrepetition errors, i.e.,TOTAL minus REP. The data were analyzed tosummarize thepupils' performance reading tests and oral reading onboth the task and to describedifferences in performance among the three groups tested. Therefore, both overall andgroup means and

143 PAGE 76

standard deviations for all pupils tested were computed for alltest scores and oral reading errorfrequenciesfor each classification In addition to described above. the error frequencies, error proportions dividing each error frequency were computedby by an appropriatedenominator. WORDSUB,ORDER, DELETE, For variables INSERT,NONSUB,NC, PC, SC,DC, NS and MS the denominator used was TOTNONREP.For variableREP, the denominator For variable SIMSPELL was TOTAL. thelenoninator was TOTSUB. Andfor variables NSSCOR, NCCOR, PCCOR, SCCOR and DCCORthe denominators respectively. were NSSUB, NC, PC, SC and DC, These proportionswere computedto permit group comparisons error types that would of be unaffected by theoverall errorrate for a certain pupil or group. For exasple, although Groups PS2 and CS had similarmeans for the frequency of NSSUBerrors (If = .18 vs. .14, respectively), this group difference was greaterwhen the proportion of considered (N = .08 NSSUB errors to TOTNONREPwas vs. .02 for Groups P32 and CS,respectively). Group differenceswere further analyzed by of variance. perforning two sets ofanalyses The first setwas performed using the scores and oral reading error reading and writing test frequencies as the dependentvariables and group (PS1, P32 and CS)as the independent variable. The second setused group again as the independentvariable but included the errors as the proportions of oral reading dependent variables. For thislatter set proportions were first of analyses,all transformed using the&rosins transformation (arcsihe of the square root ofthe proportion) to normalize their distributions. In addition to theseinferential statistical tests sizes (ES) of group differences,effect were computed to coiparethe differences, school and the comparison between each project school by subtracting themean of Group CS from that of each of Groups PS1 and PS2 and dividing deviation of Group CS. this difference bythe standard Each effect sizeconsequently standard deviation units indicates how many the mean of the projectschoolwas above (indicated by a positive effect size)or below (indicated by a negativeeffect size) the mean ofthe comparison school. As a general rule, an effect size of.5 or less is usually consideredto be indicative of effect size of .5 a negligible difference while or greateris usually consideredto difference having be indicative ofa some practical significance (seeMinium, 1978).

Table 4.19 gives themean nusber of each of the 20 oral reading errors counted different categories of forall pupilsand again separately PS1, PS2 and CS. for Groups Overall, a mean of between9 and 10 errors (TOTAL) by pupilsin each of were made the three groupsin readingthe passage of 52 words. Mosterrors were repetitionerrors (REP), followed by totalsubstitutions (TOTSUB), substitutionssimilar in spelling to errors which did not fit the target word (SIMSPELL), the sentence context (NC) and errors conform to.the preceding which did (PC) or entire sentencecontext (SC). errors were more frequent than Since SIMSPELL either PC, SC, orDC errors in spiteof the factthat SIMSPELLerrors were a subset of only TOTSUB errors whileany non-repetition error (TOTNONREP)could be classified data suggest as PC, SC, or DC), these that the pupils reliedmore on visual information processes. However, there were and top-down some notable differencesinthe types of errors made by pupils ineach of the three groups. Most noticeable is the finding thatGroup PS1 made more REP and NONSUB errorsthan theother two groups and corrected moreerrors which fit the preceding context of its sentence but not the following (PCCOR). Group PS2 made fewer the other two groups (DELETE) deletionerrors than but more multiple-sourceerrors (MS).

144 Table 4.19

Bari Oral ReadingError Frequencies

MIM IMIIMN MM WM 1 MI NM Rejaf East NNI.MMIEN M. I NI All Pupils (43) Rejaf West (Project) (18) Bungu (Project) (11) (Comparison) (14) Variable M SD SD 1. WORDSUB 3,02 2.33 3,22 2. ORDER 2.98 .37 ,02 3.091.26 .28 .15 0 0 2.71 1.38 3. REP .09 .30 5.46 2.59 - 0 6.33231 .85 0 5.451,97 .47 4.36 2.31 4. DELETE .60 1.33 .39 5. INSERT 1.24-.03 .09 .47 ,30-.71 1.29 1.08 .11 .32-.46 1.68 6. SIMSPELL 1.77 .27 .65-.49 1.62 1.61 1.07 1.64 1.94 .11 2.00 1.48 .16 1.79 7. NSSUB 1.31 .47 .93 8. NC .891.282.08 1.44 .18 .40 .11 1.94 1.062.39 .14 9. PC -.58 1.271.19 .98 1.08 -.46 2.07 10. SC 1.01.33 .19 1.73 1.09 1.09 .94 .31 1.41 .671.14 .86 ,75 11. DC -.65 .911.14 .16 -.51 1.79 .53 .11 .32 1.72 -.22 .09 .30 -.24 .29 12. SC .83 .93 1.39 .94 13. MS 1.70-.16 .64 .72 .92-.39 1.14 .93 .50 .86-.22 1.29 1.181.25 .86 .64 .63 14, NSSCOR .16 .48 .39 .70 15. NCCOR -- 0 0 0 .19 .19 .39 0 0 16. PCCOR .22 .43 .22 .30 .18 .40 .11 .64 .50 .14 .36 17. SCCOR .862.00 .18 .07 .40 .11 .14 .26 .11 .32 .36 18. DCCOR .15 0 0 0 0 0 0 .07 .27 -- 0 0 0 0 19, TOTNONREP 4.11 3.56 3.72 20, TOTAL 4.18-.39 3.55 9.58 2.42-.44 5.07 4.08 10.065.06 3.45 .17 9.002.65 -.11 9.43 330

Note,Numbers in parentheses indicatethe number of pupils ineach group.

145 PAGE 77

Table 4.20 presents oral reading error proportionswhich arein some ways more useful than simple frequenciesin comparing differences across different across groups and types of errors. By dividing error appropriate frequenciesby denominators, the resulting oralreadingerror reported in this table permit proportions comparisons thatare affected by error-rate for a given the overall group. Fromthis perspective, it isseen that overall repetition errors (REP) constituted 61% of all errors sade andword substitutions (VORDSUB) madeup 83% of all non-repetition (TOTNONREP) Evidence of sensitivity errors. to visualinforsation and bottom-up processing reading is indicated by the fact in that over half (59%) of allTOTNONREP errors were SIMSPELL errors while only 8%of TOTNONREP errorswere deletions (DELETE) and another 8% were insertions (INSERT). Thispattern of errors indicates fairly high dependenceon visual inforaation since mostsubstitutions were visually sinner to thetarget word they replaced andthe deletions and insertions low proportions of indicate careful attention tothe reading. visual stimuli of

Evidence of reading strategies using contextual informationand top-down reading strategies Ls indicated by 30% of TOTNONREPerrors showing sensitivity to preceding sentential contextual constraints (PCerrors), 22% of TOTNONREP errors confOrming to all sentential contextual constraints (SC errors)and a small 3% ,conforming to the contextual .constraintsof the entire discourse (DC errors). preceding In contrast, evidence ofinsensitivity to contextual constraintsis indicatedby thefact that 13% of TOTNONREPerrors were nonsense substitutions (NSSUB, 1.e., not real Bari words) and 33%of TOTNONREP errors were Bari words which did not conform to any sententialor discourse contextual constraints (NC errors).This means that almost half (46%) of errors made were either nonsense all words or Bariwords which did not fit the surrounding context. any of

The pattern of errnr corrections alsoprovides some insight into thetypes of information and utrategiesused bythese children to read. (1965) and Veber Both Goodman (1970b) have reportedthat children aremore likely correct errors which do not confors to tothe meaning of whatthey are reading than errors which dofitthe syntactic andsemantic passage. If constraints ofthe this were true of the childrenin the present study, expect that a higher proportion we would of NONSUB, NC and PCerrors would be corrected than SC or DC errors. In general, this patterndoes seem to hold with 36% NONSUB errors corrected of (NSSCOR), 18% of NCerrors corrected of PC errors corrected (PCCOR) (NCCOR) and 25% in contrastto only 12% of SC (SCCOR) and 0% of DC errors corrected errors corrected. (DCCOR).Thus, it appears children had at that these least somesensitivity tothe meaning of what they reading and tended to correct were errors which did not conformto the meaning of the passage more often thanerrors which did.

There were also a number ofsignificant differences which are among the three groups indicated by thesignificant F ratiosgiven in Table projectgroups 4.20. Both (Groups PS1and P52)hadahigher substitution errors (VORDSUB) percentage of word than did the Group CS as wellas a higher percentage of nonsense substitutions (NSSUB) and repetition (REP)errors. In contrast, Group CS hada higherpercentage of deletion (DELETE) (INSERT) errors than did and insertion both project groupsand GroupPS1had alower percentage of SINSPELLerrors than did both Groups PS2 CS had a higher percentage and CS. Finally, Group of noncontext(NC) errors than did either of the

147 g 7,23.0 1.29 3.1$ IAN 3.0 3.53' 4,10 3,75* 1,17 1,62 2,62 2,69 2,19 wort 2,31 1,31 2,4 2,36 1,36 2,31 2,31 2,31 2,36 1,36 1,361 2,11 2,36 2,1 2,22 2,181 " for variable 149 (Csrarlon) ng 14 14 14 14 14 14 14 14 14 14 14 14 14 10 denominators ,11 .13 ,11 .32 .07 ,31 ,21 ,27 ,01 .11 .23 .32 WORDSUS IlmOu nommr=mmi, IN 0 .41 AI11 .11 .64 .01 .46 ,11 ,3I .N ,1211 ,15 02 1120 .10 02 The 3; hert 11 11 11 11 11 11 11 10 variable mortal (PMNI) 11II 0 .11 B.11 ,41 .31 6.13II .11II .41II ,I1II ',IIII 1,11 10 1 ',II1 s Ow numerator. Air /HI ,11 111 111 .11 .33 .10 ,I0 .11 prowtism 4,11 PrIpsflom 000 .1 ,I/ fl .0 .1 1 31 1 *MkIn 13;TOTAL MIMI*, TOW Iffr (01140 1 1 ol 110040 ION 1 1,111 ,11.411 ',III Ole II,V MR.4, I Postoestims Ors 00 implimmumummm ,01 IsI, loi ,V 11 )0101ill WOO 1,1,4, lir4*OM 014su semeniamos 14 101111111101Bilell2 ,N II lIft ai1 11111 PlipPON I, Wk. mad V MON 0 4110111 to 1emillo ow VON Noah "V Hi PAGE 78

two project groups. However, in spiteof the group differences the overall profile of noted here, oral error patterns is quitesimilar groups with a greater tendency across the three to usevisual inforaation in use contextual information. readingthan to

A final perspectiveon the reading skills and strategies can be gained by comparing their of these children patternof oral reading errors other primary-school with those of Children reported in the literature. reported that most of the While Leu (1982) many studies of children's oralreading errors use their own system ofscoring which sakes comparisonsdifficult, theredoes appear to be enough similarityin the way a number of similarity and contextual studies scored graphic appropriateness to makesome Thus, the relevant comparisons possible. findings of six ofthese studies are summarized 4.21. in Table All were conducted in theU.S. except for Clay (1968, New Zealand) which was done in and all included children atthe schooling, end of their firstyear of exceptfor Hood(1975)- which examined (comparable studies second-gradechildren of third-gradechildren could not literature). befound inthe With respect to theproportion of substitution showed sole graphic errors which similarityto the textword, the sitUated in the Bari childrenare middle.of the range of theotherfive studies. proportion :of the However, the Barierrors whichconformed toat contextual least thepreceding constraints ofthe sentence seemslow when compared first-grade children. to the (Although the Bari childrenseem more similar lower proportions, of to the contextually appropriateerrors sade by the second-grade Children reported by Hood, this waybe due to whatappear difficult reading to be the more passages Hood used since shereported that they third-grade readability.) were of high

Thus, whileit mustbe admitted that thesecomparisonsshould be interpreted cautiously dueto manyuncontrolled factors (e.g., of the reading passages used the difficulty in the various studies),they do suggest that Bari children at the the end of their thirdyear of schooling were in general sensitivetothe use of contextual less informationin reading than average American children towardthe end of their firstyear of school.

150 Table 4.21

Oral ReadingError Proportions for SelectedStudies

Claye Biemillere Weberb Weberb Cohena Hoodb Variable Present (1968) (1970) (1970a) (1970b) (1974-75)(1975-76) Studyb

Grade 1 1 1 1 1 2 3 n 100 32c 19 43d 50 45 39 Graphic similarity.41 .50 .53 .. .70 .50 .59 NC -- .18 _ .38 .49 .33 PC .82. .68 (.91e).90e .62 .51 .55 72e SC -- ,64e 62e .. .22 .25 DC .., .. .12 .03

Note.To make results comparableacross studies, SC contains both SC and DCerrors and PCcon- tains PC, SC, andDC errors, This system isdifferent from that used for Tables2 and 3 where,for example, PC errors are only thosePC errors which were not also SCor DC errors. a Only substitutionerrors were judged for contextual appropriateness. b All non-repetitionerrors were judged for contextual appropriateness. cOnlychildren In "post-nonresponse" phasewere included. d Includes the 19 children of Weber(1970a), e Only syntactic appropriatenesswas considered.

151 152 PAGE 79

4.3 THIRD IMPACTEVALUATION

4.3.1 DESCRIPTION OF EVALUATION

4.3.1.1 Evaluation Design

Following the design used for th, Sari and Lotuhoevaluations in 1980 and 1981, the intention ofthe third impact evaluation 883 tocompare the reading per,mmance of Primary 1 Zande and Kuer Childrenwho hadbeen using literacy Materials since the the beginning of the 1982 schoolyear in selected "project" sehoolswith that ot Children in other "comparison"schools Who hadbeen receivinginstruction wIth other materials. The iRLhas one proj(4ct school for eachlanguage; for Zande itis the combinedprimary and middle school of Mesiya, located in Yambio. The Nuerproject isbased at Stair West Priaary schoolin the town of Nasir.

The decisionto attempt a replication of the Bari andLotuho evaluations with Zande and Nuerwas in part based on assurances we had received duringthe previous year that comparisonschools would be available in the fall of 1982. at both project sites As it turnedout this was true only three schools providing for Yambio, where vernacular instructionwere locatedwithin a fifteen mile radius of theproject school. However, latein 1982the educational authorities inNasir decidedto postpone vernacularinstructionin the designated companion schoolsfor another year. Nevertheless it was decidedto evaluate reading performanceat the Nuer project school in the hope thatdata collected theirmight be usefulin a follow-up evaluation which would examine the planned for 1983 progress of theselected 1983 Primary readers. Primary 1 pupils as

4.3.1.2Instruments

Vithone exception, thesaae type of instrumentsused forcollecting background and outcome datain thefirst and second impact maintained in this evaluation. evaluations were Cace again backgrounddata on schools--number of pupils, teachers, subjects taught, materials available--wereobtained through interviews withschoolprincipals. Early in obvious that the the testingit became pupils at both siteswere giving the same questions asked about their responses to the reasons forattending school and learning and write as they had to read during the previousyear. Sincethe pupils frequently foundthesequestionsconfusing and often produced answers thatdefied meaningful interpretation,it was decided to omit the testing. then in order tospeed up The :iteracy tests administeredat Zambia and Nasir the previous evaluation, were, as in developed at each sitein close conjunction with IRL Language Officerand the the SIL Consultant. Their format and methodof construction are described ingreater detail below.

153 PAGE 80

4.3.1.2.1 THE ZANDE TEST

The first stepin developing the literacy tests wasto determine lesson inthe IRL Primer the what students were beginning. The Literacy Officer then selected froa the letter lessonspreviously studied 12syllables which represented consonant-vowel (CV)syllables that the pupils school should theoretically in theproject be able to read aloudcorrectly. this section of the test The purpose of was to obtain an indication of thepupils' ability to recode letters into sound.Since both the "letter and at least one alternative lessons" in the IRL primers method witnessed in a Zande eaphasize comparison school the teaching of CV syllables, theuse of such a justified. testseems

The secondsubtest consisted of 10 "known"words, i.e., words whichhad beentaught inprevious lessons and 10 "unknown" (oftenreferred toas "built") words- -words which contained letters the The latter pupils had been exposedto. had often been used instories butnever analyzed and class. All of the unknown words taught in were high frequency, basicconcepts familiar to the pupils. The section containing the known wordswas included to provide a measure of the project pupils'word recognition ability with the unknown words and the subsection was intended tofurnish an index ofthe pupils' word attack skills.All of the words in both of these sectionswere selected by the IRL Language Officerand/or the SIL Consultant.

Thefinal section of the ,test consisted ofa shortpassage and comprehension questions. seven The passages were heldto a length close to the passages the that of pupils had been readinginthe IRL passage was 76 morphemes primers. The Zande long, 7 morphemes longerthan thepassage in Lesson 14, inBook 2 of theprimer, the last story they had read. (Thepupils at Masiya School were justbeginning Lesson 16 when Cowan arrived.)About 7% of thewords in the testpassage were new, unknown words, but all of these contained lettersthat the students had beentaught. Three ofthe unknown words were content words (thenouns, Mako and Bako, the who are the main characters names of two children in the story and thenoun am ("dog") and four were grammatical morphemes ka "in order to" and uko "topour from" restricted to the sense of "many small objects pouring froma container", ti "to fall" in the sense of a singleobject and ora "to run"). One ofthe comprehension questions, number six, couldnot be answered by quoting a sentence a sentence in the story. or part of However, the answerwastransparently obvious from the context. Pilottesting with 6 children revealed thetest were of an appropriate that all subsections of levelof difficulty. The completeZande literacy test with Englishglosses is shown in Appendix 4.A. 4.3.1.2.2 THE MUER TEST

The SIL Consultant, Marian Farquhar, had, with the encouragement of SIL Linguistic Consultant Senior JocelynClavinger, attemptedto group receiving literacy instruction the pupils in Nasir according toability. most advanced group consisted The first and of 21 P2 pupils whowere at the time beginning Book 4 of the IRL Nuer primer series. Thesecond, intermediategroup was comprised of 17 P2 pupils whowere working on Lesson 20 of Book 2.The third group encompassed roughly 100 beginningP1 children who had reading beginning IRL primer no books and were lessons taughtfrom the blackboard. Within this last group were repeatersfrom P2 who had had no previousreading instruction

154 PAGE 81

and 16 pupils who had been promoted to P2 but had sade such littleprogress in reading during the previous year that itwas feltthat theyshould repeat beginning instruction.

Since there was no possibility of obtaining comparison groupssimilar to thosefound in Iambic), Itwasdecided tocollect data onthe reading performance of as many P1 pupilsas possible, in the hope that this might be useful for a follow-up study in 1983. However, an obligation was also feltto honor the SIL Consultant's request thatthe tenting provide practical feedback of immediate use to the program at Nasir --specifically that it providea basis for determining whether thegroupings truly reflected distinctlydifferent levels of ability. Accordingly, three different testswere designed--one for each of the aforementionedgroups.

The test constructed for the highest group dispensedwiththe CV syllable subsection, since virtually every alphabetletter in every position is covered by the time the pupils reach Book 4 As can be seen from Appendix 4.8,the known and unknown words aremono- and polysyllabic sometimes polymorphemic and, again all commonconcepts the children would know. The passage used in this test was an altered version ofa story which theNuer Language Officer had written for Book 3. It is a little over 200 wordsand has a clearly defined narrative structure.The longest polysyllabic wordsin the story are all names of places in the immediate vicinity .of Nasir,whichthe students should be familiar with. Three of the comprehension questions- -numbers three, four and six- -requiredthepupil to makeinferencesfromthe story. Acceptable answers to questions three and fourcould be inferredby having understood certain paragraphs in the story, but question six requiredthe reader to have understood the entire story and recognized how seriousthe main character is about education and thathisheadmaster wishes to reward him for this.

The test created for the second, intermediate group had a shorterpassage, comparable in length to the stories the students were readingin Book 2. It 's the same length as the Zande literacy test. The knownand unknown words _Are largely monosyllabic. There are sevenquestions forthe comprehension passage.

The test designed for thelower level students containedaCV syllables section, but since the pupils had beentaught only three consonant lettersand two vowel letters atthat time, they were testedonjust six syllables using combinations of those symbols. All of the known andunknown wordswere monomorphemic and monosyllabic. The comprehensionpassage was64 morphemes long; it was followed by seven comprehension questions, none of whichrequired inferencing. The three tests developed for thepupils at Nasir West are shown in Appendices 4.0 through 4.E.

4.3.1.3Data Collection

Background data on teachers, pupils and availabilityof materialswere gathered through interviews with headmastersandteachersusing thesame

155 PAGE 82

questionnaire used in the previous evaluations. There certain points as reported were discrepancieson by differentindividuals, so accuracy of the that thecomplete information displayed inthese tables cannot best these data provide be assured. At a general picture of theconditionsfound atthe schools visited duringthis evaluation.No attempt was made to background data on teachers collect at Nasir Vest, sincethis was impact evaluation. done in the second It is worth noting, however, that a different teachers is shown for Nasir number of Vest this year than forthe previous year. is difficult to explain,but This last year's data should be quite accurste sincethey were supplied by COnsultant, Marian Farquhar, the SIL who is extremely meticulousin such matters. in the previousyear, As accurate data on attendancewas available only for the project classes at NasirVest.

Process data were collectedat three of the four Zande West by tape recording schools and at Nasir and takingnotes while observing lesson. a vernacular literacy

Essentiallythe same interview proceduresemployed inthe past evaluations wereused with the two literacy teststo collect outcome data. At each school thetesting took place in a room set apart fromthe areas where teaching was goingon. The pupils were testedindividually by the Language Officer (in Nasir by MarianFarquhar) while an audio recording was made ofthe entire interaction. Aa each pupil read thetest aloud, the syllables and words incorrectly readwere recorded on a specially prepared tally sheet. An attempt Was also made tokeep a written record of on the passage, the oral readingerrors made but due to the variationin reading speeds, this accurate at those was not as taken for the firstthree sections of the pupil had finished reading test. When the the passage, the LanguageOfficer asked each of the comprehension questions andrecorded incorrectanswers. and evaluator The Language Officer then conferred to determine if our respective reading errors were identical. tallies of oral Alldiscrepancies were checked the tape of that section. by replaying In order to simulate theconditions under which the P1 pupils had done mostof their reading at the Zande comparison schoolsand at Nasir Vest and to avoid potential confusions that mightarise because two lettersin theprinted test, g and a, differed in type from thewritten versions they were used to,some or all of the subsections of put on the blackboard. the P1 test were Cowan would then pointto the items to be read Language Officer or the SIL and the Consultant would note andrecord any errors.

4.3.2RESULTS

4.3.2.1 Background Data

Characteristics of the five schoolsincluded in the third impact are described below with evaluation a suamary of background datapresented in Table 4.22.

156 Table 4.22

Background Data

School

Masiya Maingbangaru Susa Singbi Variable (lande Project) Nasir West (Zande Comparison)(Zande Comparison) (Zande Comparison)(Nuer Project)

--reiera TeacherData

Number of Teachers 12 8 3 8 8 Above teachers with regional mother tongue 12 (100%) 8 (100%) 3 (100%) 8 (100%) Trained Teachers 8 (671) 7 (87.9) 3 (100%) 6 (751)) Experience (years)

Mean 14.5 28.67 26 26.25

Median 17.5 30.5 28 18

Range 1/3- 30 6 - 50 Iw~OftrlaN20 - 30 4 - 50

( coat d)

157

158 Table 4.22 (cont'd)

Background Data

School

Masiya Maingbangaru Susa (Zande Project School) Singbi Nasir West (Zande Comparison)(Zande Comparison) (Zande Comparison)(Nuer Project)

General Pupil Information

Enrollment:

P1 60 80 38 80 2331 P2 27 18 37 35 1381 P3 36 27 25 25 55 P4 a.11 14 15 17 40 P5 18 15 ... 5 40 P6 36 MO -- 13 40 Total: 175 154 104 175 546

1 These represent official enrollment figuresfor Naslr West but attendancefluctuates considerably every day.As few as 60 to 80P1s and between 50 to 60 P2smay attend after breakfast. If this is takeninto consideration, total enrollment is reduced con3iderablywith a conservative estimate being245.

159 160 PAGE 83

4.3.2.1.1ZANDE SCHOOLS

The followingfour primary schools in and around Yambiowerevisited for the purpose of exasiningP1 pupils and observing literacy classes. All of the schools had been in sessionsince July.

4.3.2.1.1.1Masiya

Masiya school is a combined primary and middle school. It islocated on the edge of Yambio,a town with a population at .twelve thousand. somewhere betweenten and The capital of WesternEquatorial number of government Yambio, houses a large departments, among them, the Police (bothdistrict and provincialoffices), Labour,Health, Wildlife and Tourism, Co-operatives, Social Welfare, Agriculture,Forestry, Commerce and school is Veterinary Medicine. The located near several villageswhich form the outskirts It is abouc 200 yards off of Yambio. the main road leadinginto the center ot town.

There are five buildingson the school compound, one of which is usedas at teacher's room and a storeroomfor equipment. TWo of the other fourbuildings have mud walls and thatchedroofs. The first of these has is used for one classroom which the middle school pupilsand a small room that house SWIRL shortwave formerly used to radio but ispresently the school caretaker. living quartersof the The middle school classroomhas a large fixed alick.blackboard permanentlyto.onewall. Benches and writing surfaces constructedout of split logs have been supportedby forked tree crotches securely in the hard packed anchored earth. The ud walls aroundthe class are raised to a height of about 3feet. The other building withmud walls is locatedat the farend of the compound. It is about 100 feetlong andcontains two classrooms at either end, each with a fixed blackboard the and one small room in center which isused 'solelyforstoring equipment. classrooms are simply Theother two shelters--thatched roofssupported thatched wall dividing by poles--witha this enclosure into twoteaching areas. on the ground, often on bricks The pupils sit and are either taught froma blackboard ',misting on a tripod orfrom their books. A few classes still neet under large shade trees, but constructionon yet another shelter-type itwould appear that classroom is underway and the schooliscontinuingto expand facilities. itsclassroom

Of the four Zande schools visited, Masiyais the largest and best both in terms of teaching equipped materials and number ofteachers. vernacular teething materials. Only Masiya had

4.3.2.1.1.2 Maingbangaru

Maingbangaru Primary School is located aboutseven Yambio on the miles northwest of road to Nzara, avillage of reportedly 8000 school islocated in a spacious, inhabitants. The open areaon theside of services the villages the road. It between Nzara andYambio, providing pupils in P1 instruction for 154 throughP5 (P6 isplanning for 1983). shelter-type classrooms, There are three one of which (the P5 class)hasbenches and writing surfaces made of splitlogs. A thatched-roof shelterfor the P3class was under construction. The other classes were held under large trees where log benches have occasionallybeen provided. P1 and P2 classeswere receiving literacy training and althoughboth headmasters claimed that the studentshad

161 31111:1x 'II %Mb1115911.1I: I alepili 11E111 yiIff pis I 1 !Pi I/ Pal %IA 11 Sh 410111 NT41 6;10110 DS211Ala! is si 11° ei11 :digNI I IIIT ?gill, i lad is %OVA1 1. TT it1014 lail I II/ 3111 PAGE 85

Ethiopian border. It is fairly large, station, a military possessing a number ofshops, a police cantonment, a mosque,a Lutheran mission, churches, school and two primary schools. a high Boats occasionallycone up the Malakal and stop atNasir. river from

The school consisted of three buildings, all mortar walls supporting of which tad mudor mud and tin roofs. The firstof these, a "U" shaped housed classrooms, the headmaster's complex, room. room, a storage room andateacher's The other two buildingswere one room structures feet. Sitting on the dirt approximately 80 by 20 floor oron stacks ofbricks they them, as many as 120 brought with Children occupied theseone-room schoolhouses. Blaokboards wre hungon the wall at the front of the class butremoved at the end of the school day.Some classes of a building. were taught under treesor in the shade

As Table 4.23 shows,six subjects were offered in grades P1 through P6. In addition to Simon Kuoon,the IRL Language Officer whom teach only and Marian Farquhar,both of literacy, the school hada staff of eight theseinstruct different teachers. All of subjects at differentgrade arrived in early November levels. WhenCowan the school had beenin session due to various kinds of only about 53 days, interruptions ranging frompolitical disturbances.to planting and harvestingholidays.

Attendance at theliteracy classes had increasedconsiderably since 1981. The official rostershowed 150 boys and 83 girls in Pl. Average attendance before was 100 to 120; afterbreakfast this dropped to between 60 and 80. The sane pattern was seen inP2; of the 138 attend before registered children about70 would breakfast, between 50and 60 after breakfast. problem of achieving Facedwith the effective instruction for such literacy teachers had large numbers, the two grouped the pupilsin the manner described They had just received earlier. some added assistance in the form of four observing classes to learn men who were the methodology practicedat the school.These men were employed by Mt. MichaelO'Brien, a Community DevelopmentSpecialist from the Lutheran WorldFederation.Mt. O'Brien is interested in using literacyas the main vehicle of hiscomaunity development eager to obtain programs aaong the Nuer andwas training in literacymethodology for his Vest. assistants at Nasir These assistants had almostreached thepoint where they take some classes at the were ready to school. Itwas Marian Farquhar's intention up the existing groups into to break smaller, more evenlygrouped classes and to four of these to thenew teachers. assign

4.3.2.2Process Data

4.3.2.2.1 THE ZANDE SCHOOLS

Different lessons were observed atMasiya, Maingbangaru and letter lesson at Masiya Singbi. The was a very close repetition ofwhat has already been described in the secondimpact evaluation. The writing lessonat Masiya was also virtuallyidentical to what had been observedthe previous year. The teacher would writeselected letters onthe blackboard and then ask children to identify them. the Having done this, thechildrenwould first watch

163 Table 4.23

Subject, Curricula andMaterials

Grades Receiving Instruction in Language of Minutes Pupil's Subject this Subject Teacher's Instruction per week Materials Materials

Masiya School (ZandeProject)

English PI- P6 Zande 240 None . Math PI - P6 Zande 240 None . History P3, P5, P6 Zande 240 None .. Geography P3, P5, P6 Zande 240 None . Science P3, P5, P6 Zande 240 None .. Religion P1 P6 Zande 240 None Arts . P1 P6 Zande 240 None .. Arabic PI - P6 Zande 240 None .. Vernacular P1, P2 Zande 240 IRL Primers Teacher's Guide

Maingbangaru School(Zande Comparison)

English PI - P4 Zande 240 None 1 copy Math 151 - P6 Zande 240 None 1 copy History P4, P5 Zande 240 None 1 copy Geography P4, P5 Zande 240 None 1 copy Science P6 Zande 240 None None Religion P1- P6 Zande 240 None 1 copy Arabic P4, P5 Zande 240 None 1 copy Vernacular PI Zande 240 None 1 copy

164 (cont'd) 165 Table 4.23 (cont'd)

Subject, Curricula andMaterials

Grades Receiving Instruction In Language of Minutes Pupil's Subject Teacher's this Subject Instruction per week Materials Materials

Susa School (Zande Comparison)

English PI - P4 Zande 240 None None

Math PI - P4 Zande 240 None None

Geography P3, P4 Zande 240 None None

Science P3, P4 Zande 240 None None

Religion PI - P4 lande 240 None None

Vernacular PI - P4 Zande 240 None None

Singbl School (Zande Comparison)

English PI - P6 lande 240 None 1 copy Math PI - P6 Zande 240 None 1 copy

History P3 - P6 Zande 240 None 1 copy

Geography P3- P6 Zande 240 None 1 copy

clence P3 - P6 Zande 240 None 1 copy

Religion PI P6 lande 240 None 1 copy Vernacular PI Zande 240 None 1 copy

166

167 PAGE 86

andthentrace in the air the purposelyexaggerated movementsofthe instructor's hand as he wrotethe letter. Next, the pupils would letter in their books, write the or, more frequently, in the dirt (usuallyon lines they had drawn to simulateruled paper.) The procedure was then entire words and repeated for the finally, short sentences usingfamiliar words letters. with the same Theteacher is supposed to go aroundandcheck the periodically. pupils' work These check-ups are quiteimportant, since aany of the have serious problems with students mirror image letters, i.e., b, d, 2, and letters distinguished by only a single stroke, e.g., r, n, a. Quite a few children were seento be writing vowels like e and a backwards and one child actuallywriting letters was rotated at a90-degree anglefrom their orietation. normal It appearedthat theteacher had either decided unable to that he was provide sufficient monitoringfor allpupils, or had policy that he would adopted the only help those students hefeltcould be helped. pupils who knew how to Some makethe letters correctlyoccasionally who were having difficulty. helped those This seemed to ae at the time to be worth encouraging. a policy

At Naingbangaru, an activitywasobserved that may have lesson in recognizing syllables been a review or perhaps a rather lengthyreview followed by the introduction ofsome new letters. Thislesson was taught to class of P1 and P2 pupils, a combined all of whom presumably hadno books. wrote about a dozen CV syllabAas The teacher on the blackboard.He then pointed syllable and to a asked for a volunteer toread it. Pupils who volunteered but were unable to read a syllablewere told tosit down and another pupil called upon. was When a'correct readingwas produced, the teacher would the entire class chant then have the syllable at leasta half dozen times. The lesson would proceed this way until all of thesyllables (there were about 12or 15) had beenread and repeated. Then theteacher went back andbegan calling individual pupils and having thenread a syllable. able to read only Often the student would be the vowels. When this happened, theteacher would attempt to p-' tfr pupil to sound out theinitial consonant.If the pupil was unable to, the Le. .111' wc.old call on otherstudents until he finallygot a correct rendition. The teacher would thenbreak down theCV syllable into its component pPrts and have the class repeat then. In.doing so, he would commit o of the uasic errors that proponents of phonics have attemptedto correct fcr years the insertion of a schwa aftera consonant. Thus a CV sequence like /ra/ tould read /r a/. The consequences of thisprocedure were evident in the perfcAmance of some of the students in thesyllable subsection of the mading trJc. The entire lessonlasted about 40 minutes. No reading of meaningfu' aaterial was observedduring the lesson.

che first and second impact evaluatiot,4; it was argued thatthe high priority placed onoral repetition of syllablesand words does littleto promote real learning and in fact may be detrimental when tha sentencesin the primers recyclewords frequently. Nowhere was this sore evilant %nenin an English lesson forP1 pupils which took place Singbi primary teacher was using a chart of *school. The illustrations to prompt productionof sentences with place expressions in response to "where" questions. He would point to a picture of a man undera tree and ask, "Where isthe aan?" and the entire classwould thenchorusthe appropKate answer, which memorized. The fact they had long ago that this repetitious chantingwas inculcating language behavior that could not be extended to new contextswas quickly demonstrated when the evaluator asked the pupils some of thesame kind of questions using

168 PAGE 87

vocabulary they all knew. For example, having pointed to theirteacher and indicated that he wasstanding under thetree, the children were asked is "Where the teacher?" All ofthe childrencane back with a word-for-word repetition of the sentence they had been memorizingwiththe chart- -"The is standing under man the tree." Identical responses wereobtained to questions like "Lhere am I?Where are you?"

4.3.2.2.2 THE NUER SCHOOL

ttNasir an interesting twist to ending alesson was observed which de.ste.rves reporting, since it effectively engaged the attentionof alarge class in some consciouslearning.The Language Officer, teaching one of the Simon Kuoon, Was primer stories from the blackboard. He wrote the two paragraph length story on the blackboard,having first divided the classinto two groups. After allowingthe pupils some time to read he asked the story silently, the comprehension questions,which were answered by vlunteers. individual Next, one group of children (halfof the class on one side room) read one sentence of of the the story in chorusas the Language Officer pointed to each successive word in the sentence. The next sentence was the other group and this procedure chorused by was repeated until all of the sentenceshad been read. Individual reading followed. When a child faltered, theLanguage Cfficer wouldignore the other pupils' cries to be called on andhelp the reader analite andsound out the word sothat he almost alwaysgot it. The Language Officer seemed to seekoutboth strong and relatively for this activity. weak readers This sentence-by-sentencereading consumed a fair amount or time, Ls the Language Officer apparently wanted to giveas many children as posslr,I., an opportunity to read. Finally pupils were calledon to entire story. read the

The lesson concluded with a wordreading activity which Cowan originated with Simon run. assumed He wouldask: "Now who can showme the word ?"Fifty hands shot into the airas the entireclass volunteered. The pupil selected would then go to the board and point toa word. If it WS3 the correct word, the LanguageOfficer erased it, if not, the child went to his seat and another pupil back was called on. There waslots of good natured chidingby theteacherand the entireclass of those students whomade incorrect choices, all of which is perfectly acceptableamongthe Nuer. The gaae went on this way until all of the words were erased from the board.This activity seemed to have two things to recommend it: it enlivened the class at the end of a long lesson, while providing the teacherwith a check on the word reading ability of a groatmany pupils.

4.3.2.3 Outcome Data

4.3.2.3.1 THE ZANDE LITERACY TEST

Forseveral reasons it was not possible to obtain equal-sizedsamples across schools. Only 16 P1 pupils were testedat the projectschool.* The Language Officer and the twoteachers at Nasiyastated that there fact, additional were, in P1 pupilsreceiving literacy training, but these their opinion, "underaged," were, in and had therefore not madeas munh progress as the

169 PAGE 88

children theyselected for testing.Therefore, only those pupils that the Language Officer and theteachers declared were capable of all weretested and so reading anything at there is no way of knowingwhether this sample biased. The group of P1 children was whotook the testwere divided into two categories:children deemed aature enough to take their primershome with them after school (P1A pupils)and others who were not granted (P1B pupils). this privilege

Only 10 PI children turnedup at Susa Primary School during there. Cowan's visits The explanation given forthe poor attendanceon both days the children had a tendency to was that skip school after aheavy termites. rain tohunt There were, however,a nuaber of P2 pupils at the school curiosity to see how well and out of they would peeforaonthe test we had developed, eight of them were tested.

Maingbangaru producedthe largest pool ofP1 pupils. In the course of three visits wetested 24 pupils,five of whom were previous year. repeaters from the Singbi wasthe last school visitedand it school officials, was there that the no doubt concerned about how thepossible low performance of their P1 children might reflectupon the reputation of the school, deception. attempted a On oursecond visit, the LanguageOfficer and the first child Cowan noticed that sent to ua seemedverybig for a P1 pupil. read with a facility that far Furthermore, he outstripped all of the eight P1 pupils we had exaained on the precious day.Suspecting that we might on our have a P2 or P3 pupil hands, the Language Officeasked the boyto fetch his book after he had finished class writing being tested. The child walked directly class. to the P3 When he was half wayacross the compound a number of children who were waiting the other to be tested shouted outtohia in principal said we were supposed Zande: "The to gosit In P1 today!"We stopped the testing, went to immediately the P1 class and compileda new list of children to be tested. Nevertheless, six more P2 pupilswereslipped inthat day. They were all discoveredby the same ruse or by elicited their asking trick questions which true class. The attempteddeception thus reduced number of P1 the actual pupils at Singbi from 22to 15. The breakdown by grade total number of Zande pupils of the tested at all schoolswastherefore as follows: 16 P1 at Masiya, 24P1 at Maingbangaru, 10 P1 and 8 P2 at Susa and 15P1 and 7 P2 at Singbi (Totaln = 80).

An inspection of theraw data revealed that the performance ofsome of the P2 children at Susawas indistinguishable fromtheir P1 counterparts. Moreover, the performance ofPis at all comparison schools that one might actually was often so poor, entertain the hypothesis thatthe project children had outperformed all of the childrentested at the comparison this, analyses of variance schools. To test (ANOVAs) were runusing the children tested (80) with total numberof group as theindependent variable and sections of the test as the scores on all' dependent variables (seeTable 4.24). Significant differences werefound only for the Nonsense Syllablesand KnownWords sections ofthe literacy test2 < .01. Posthoc Fisher LSDtests revealed significant differencesonly between Masiya and all otherschools onboth graainTal-lif the 16 children testedbelonged to the"underaged" group. Therewere 8 P1A5 and7 P183. Hence the t-testsused tocomparethe performance of these twogroups show only 13 degrees of freedom.

170 Table 4.24

Analyses of Variance for P1 Pupils' Performanceon Zande Literacy Test

Section Source of Variation Sums of Squves df MS

Nonsense Between Groups 619.2 3 206.04 17.88" Syllables Within Groups (Error) 702.72 61 11.52

Known Between Groups 441.03 3 147.01 16.76** Words Within Groups (Error) 534.97 61 8.77

Unknown Between Groups 1051.68 3 350.56 10.40** Words Within Groups (Error) 635.01 61 10.41

Reading Between Groups 112.33 3 37.61 8.17" Comprehension Within Groups (Error) 280.6 61 4.6

** < .01

171 co CO 172 Of PAGE 89

parts of thetest, 2 < .01. Thus, when all P2pupils were left in the comparison school sample, the project P1 pupils outperformed the compariaon schools the children in on only two sections of the literacytest. Quite a different picture emerges when the P2 children tested attwo of the comparison schools are removed from the sample (n = 65).A3 Table 4.25 shows, significant differences were found on all sections of the literacytest. Onthe Nonsense Syllables sectionthe project pupils did better than significantly the three comparison schools,2 < .001 and Mainbangaru school pupils were significantlybetter than Susa and Singbipupils, 2 < .01. Virtually the same picturewas seen on the Known Words section The project pupils significantly of the test. better than the comparison pupils (2 < .001 for Misiyavs. Susa and Singbi; 2 < .02 for Masiya vs. )Iaingbangaru), and, again, Maingbangaru pupils were significantly better thanthe P1 pupils at the other two schools, 2 < .001.However, there was no significant difference theperformance or the in project pupils from thatof the P1 Maingbangaru on the Unknown pupilsat Words section. Theperformance of both project school and the Maingbangaru pupils was significantlybetter the pupils at the than that of other two schools, 2 < .001. On the Reading Comprehension section the project pupils performed significantly betterthan pupils at all other schools (2 < .05 for Masiya vs.Naingbangaru, 2 < .001 for Masiya vs. Susa, Singbi). Again, Maingbanbaru pupils performed significantly betterthan the Children tested at Susaand Singbi, 2 < .01.

Given the Beall size of the groups andthe factthat the selection of pupilstested wasnottruly random, the ANOVAs are bestinterpreted for descriptive purposes only. Table 4.25 is perhapsmore useful for comparison purposes. Looking at the means of each group, we cansee the project pupils clearly surpass the comparison pupilson all sections except number Words, where the pupils at Unknown Maingbangaru dobetter than the project pupils. A convenient explanation of thismight invoke the difference citethe high scores of in sample sizes and the five students whowere P1 Maingbangaru. repeatersat However, this solution is perhapsto facile, since there were eight non-repeaters who hashigher or equivalentscores. the oral reading errors made An examination of by both groups providesaneqrally compelling explanation.

The Maingbangaru pupils oral reading errors tended tohave a close similarityto the graphemic, i.e. visual shape, of thewords they attempting to read. were For example, they said irofor ira, bai for kid, irifor bibiri, ari for Dari, same for suma, !map for This is exactly what pzengu, la forans2, etc. one would expect with pupils whohave grasped the basic concept of word attack, i.e., convert letters intounits of soundand amalgamate those into higher units which can then be recognized morphemes. as words and The project pupils, on theother hand, tended to substitute that had appeared with high words frequency countof all of themorphemes used in Lessons 1-15 in the Zande primer, seeAppendix 4.F). For example, Unknown Words section, for on the meme (goat) Masiya pupils readmama (leopard) which had appeared 17 times in theirprimers; for tita (grandfather) they (hoe), which had appeared said sila 10 times; for Dari-(WOg)they name which had appeared 33 read Dabira, a male times; for nzeme (fat), theysubstituted nzengu (market) which had appeared33 times in their primers.

173 Table 4.25

Zande Literacy Test Results

School

Masiya Maingbangaru Susa Singbi (Project) (Comparison) (Comparison) (Comparison) nber of Pupils Tested 16 24 10 15

.84 3.67 1 2.24 Range 0 - 10 0 - 10 0 - 1 0 - 9 Reading Comprehension (7) Mean 3.7 2.16 o .63 Standard Deviation 2.14 2.56 0 1.83 Range 1 - 7 0 - 7 0 0 - 7 ans mt underscored by the same solid line differedsignificantly( e <.01)

; in parentheses indicate the maximumscore for each section of the test. ntity of the figures shown in sections2 snd 3 for Singbi school is purely iraphical error. coincidental and not the result

'4 175 co FAGS 90

Ito sae SOO of awhatitellses vers mode inthe reedingoomprehension r. Oa oftrootoro la the test passage were named and il=9"11= :!4414611111 711:ire :!!!!..td tEhrea

= rig Is. "IwoamiesiZttorsaj: Iderlirmi:mtps tir COMO (Ibb hod Jost at beam latredesed la Lesson16). Meld ambstitete verdethat were erten absurd SOMA ar Ole OA mere graphsalsolly in the SMO Wee I. e.g.. gel (tersites) tOmiolloss) to them as smears to-the 91111 WINN eIgtwim Question ewes mat frequestly appearedin in; IMMMDA--- Miele adbaltubsOOur a new wsrd la the 0.41.. Okapi essprehension mime =Om) la theslaws, (when) f:::714 rzje111101,.00/ IMMO, trIMI, "When were running, p amil bass geed in the primer 30 hatlPst Masa lesssam4 times in

lbstgos at eat rsalliss error ere memouly oft by tinproject pupils assaMploggests Met tamaro sok as asn amparlasa pions. boo 4aping word attack skillsas rapidly *sass 4 I. primers recyclewords so =4111'OM olikerstolih000 devabsped a guessing strategy basedon a 11611011611110 Se Imo vord te sew word mad their apparent800s garfOissmas111 dna astsalip be a dememstratien 11111116D et their ability to ISM Mee Me sowld tosany Uwe. Since the project MON IIISOPSOMMaidMO at ask a 1110 rate, *tile IMO op MOW. lt the development of word Ss lave set sorprialm,&Nat the Sande project =sISMS MOM Wand Isthot mottos of literacytest which taps oliattletonoo sot aldamillar verde without the benefit ofcontest. Iffssitteig alds4Millaisasp is the mint pupils,demonstrated ability to amatast WM* tolottlystsown abut umfmaillar words. =eassalap se Mott Comparison ea assallas set sagaindividuml grapheme that MOON lea Mid. ad theattire saint et a sestina ON lift flash 114 Us Wass la forgotten mars at Mesiya frequentlyused content in Sedbill 1,1 Oita Who Obey sass te word they did not knowand 0118 nratgartatasambease be galmapessible clue tothe 'meaning WINO VOW ter& laseal reatilagovers fatlas project pupils made =smid mootto goodmos vitals timi ag ISOWSIIII WSW asateabe aim they appeared, WON 5111ast, rambling sort often the ININSIMIIIMIS of the problemverde e.g., ad (sold) forkoho 111 (1) Nal& ONO lerlet (WO la (2), -a-ich resultin-the 111.411111diai sollstss16) (a) vita asteriM.

( $ Id eft Imo aghe ahem ma Mho Itor Ohm p Ia ardor bebey bases amd pumpkin easy mist beMgOsoos NS pumphia. lea rad est

IlliamirWmboth fres the last half et the She tot gosooso, Watt, latroductorysentence of abridged fir presemtationpurposes.Thepast swam a Is awe taasatad la treat et tap verb since it has alreadyappeared Win NI Mod Mt) la %besestemse.

176 PAGE 91

b. 1 ki ndu ku ngbe abapu nah gi *They thengo in order to buy beans and sold *"They went to buy beansand sold."

(2) a. i ki ndu ku nzenguyo ku ngbe abapu They then went tomarket there tobuy beans "They went to the marketto buy beans."

b. 1 ki ndu ku nzengu yo ku ndu abapu They then went to marketthereo went beans *"They then went to the marketto went beans."

Table 4.26 shows the percentage of the P1 pupils'oralreading which made syntactic orsemantic good aanse. The perceatagesarebased onthe total number of recordederrors and would be misleading without the number of pupils who some indication of were unable to read anythingon various sections of the test and hence producedno errors.The bottom half of Table 4.26 this additional information. provides

T tests were computed comparing theP1A project pupils' performanceon the literacy test with that of the P18 pupils.On each section the P1Apupils did significantlybetter (Nonsense Syllables: t (13) = 3.62, 2 < .01; Known Words: t (13) = 4.62, 2 < .001; UnknownWords: t (13) = 5.45, 2< .001; Reading Comprehension: t (13) = 3.01, p< ^1).Obviously these results should be interpreted with caution, since the classification cited by the project teachers "more mature," as the basis for deciding which pupilsshould be allowed to take their primers home,mightconceal factors which were the source of true the P1A's superior performance, e.g. only the older best readers children or were assigned to thisgroup. Itappeared that"more mature designated those children who the teachers felt could be trustednot to lose or damage their primers. Ifthis is true, theresults of the t-tests thus would tend to support the hypothesis made in the second impact evaluationthat children who are given material to readat home will read more frequentlyand hence develop proficiency fasterthan children who are notgrantedthis opportunity.

Theoverall results are notparticularly surprising. One wouldexpect pupils who have been using structured reading materials to perform betteron a test that taps skills practiced in those materials than children whohave not had the benefit of any materials. Still, we cannot exclude the possibilityof enviromental differences being atleast partiallyresponsiblefor the superiorperformance of the projectchildren. Given thelarge numberof government offices in Taablo, it wouldnot bestartling tofind that parents of the the project childrenare bettereducated and more, supportive of reading thanmost of the parents of thecomparison pupils.Onewould also anticipate the possibility of more literature in Zande being in the homes of the project pupils than in the hones ofthe comparison pupils.

The elusive potential contribution of the family and unidentifiedcommunity institutions to the acquisition of literacyin primary schoolsurfaced once again in the course of this evaluation. At Singbi, a number of olderchildren wereseen sitting around listening to one childread aloud from a small pamphlet which containedscripture passages in Zande. The children who were

177 Table4.26 Percent of Oral Reading Errors That Made Syntacticand/or Semantic Sense hool Made Sense Made No Sense (Project) 72% 28% ngaru (Comparison) 28% 72% (Comparison) 27% 73% (Comparison) 43% 47%

Percent of Students Who Read Nothing On Sections ofthe Literacy Test

Nonsense Known Unknown Syllables Words Words Story

di* NM, 0.11. igaru . 111 4% 4% 29%

WIND 60% 80% 100%

IND 54% 74% 93%

179 rn a PAGE 92

listening and the reader were all P2 andP3s. The reader appeared fluent that itwas decided to spot-test him to be so and several of his P2comrades, using the P1 test andsome questions about the passage he had just read. Four P2 pupils were testedin this manner and all performed very well. Similarly, the P25 whom the Singbischool officials attempted to pass off as P1s,were good readers, something thatcould not be said for all of theP2 children tested at Susa, sone ofwhom were virtually illiterate. one of the P1 Andyet, all except pupils we tested at Singbidisplayed very poor reading skills. It was not clear how,with so little time left in the schoolyear, the Singbi P1 children could makesuch giant strides to attain the level.ofcompetence we were witnessing '11 the P2pupils, who had school; the gol had only about 6more months of theirrespective abilities simply obvious explanatLon seemed toogreat. An is thatthereis sorethan instruction in some communities. one source of reading With the help of the Singbiheadmaster, the P2 andP3 Children werequestioned in an attempt read at Church, or at to determine whetherthey church-associated functionsand how many ofthem had pamphletslike the one we hadseen. Ofthe thirty to forty P2 andP3 children present that day,only about six admitted and none said that they to having any books athome read anywhere'ekcept atschool. which this informatton Due to the manner in was solicited, it should beconsidered highly suspect. 4.3.2.3.2 THE MUER LITERACYTEST

The results of the three literacy testsadministered atNasir West school are shown in Table 4.27. The P1 pupils did better on allsectionsof the literacy testthan the P2 pupils whoare in the Only on Known Words first yearreading class. was teir performancesignificantly better than beginning readers, t (31) = the P2 2.85, 2 < .01. Although there is' no strictly valid way of comwingthe P1 Muer pupils' performance with that of the P1 Zandepupil3, sitioe equivalence between thetwo literacytests isnot demonstrable, a comparisonbased on the test Nuer project scores would suggest that the pupils are better readersthan the Singbi, but Zande pupils at Susaand not as good as thepupils at!lesiva and Maingbangaru. worth noting is What is that the difference betweenthe reading skills of and Muer Children did not the Zande appear to be as pronouncedas it had appeared during the second impact evaluation.

Table 4.28 showsthe Pearson product-moment correlationscomputed to determine whether arelationship existed between attendance at schooland performance on the literacytests. Most of then are close the large negative correlation to zero except for obtained for tha P1 pupilson section and the .56 obtained the Known Words for the P2intermediate pupilson Comprehension section, which the Reading say be specious because ofthe small size of 1,:hat group. That, once againno apparent relationship performance on between attendance and the test turned upis probably due Consultant was responsible to the fact that the SIL for selecting the pupils tested. A more random selection and largersamples might well have produced evidence ofa positive relationship.The bottom half of Table 4.28 providesa more complete picture of the attendance of thepupils tested.

Only about 5% of the oralreading errors made by and Unknown Words section the P1s on the Known Words were not real wordsor morphemes in Muer. percent ofthe errors made Thirteen by the 123 beginningreading on the section were of this type. Known Words Project pupils were oftenunable to read the final

180 Table4,27

Nuer Literacy Test Results

Test

Known 1. Advanced Test Unknown Words (10)° Words (10) Comprehension(6) P2 Pupils(n = 15)

Mean 8.46 6.43 Standard Deviation 3.83 1.83 Range 3.19 2.42 1 10 1- 10 1 - 6

Known 2. Intermediate Test Unknown Words (10) Words (10) Comyehension (7) P2 Pupils (n= 7)

Mean 8.43 4.64 Standard DevIE:,% 5.29 1.56 Range 2.34 1.19 5 - 0 10 1.5 - 8 3 - 7

Nonsense X nen 3. P1 Test. Unknown spai_its. (6) Words. (10) Words (10) Comprehension (7) A.P2 Pupils Beginr0,4 Reading (n= 13)

Mean 1.53 1.95 ,25 Standard Deviatbn 1.53 1.66 2.57 1,33 Range 2.04 0 - 6 0 - 9 0 - 5 0 - 1

B.P1 Pupils (n=20)

Mean 2.58 4.8 .4 Standard Deviation 1,91 1,70 2.84 .97 2,02 Range 0 - 6 0-10 0 - 4 0- 6.5 a Numbers in parenthesesindicate the maximum score for each section ofthe test.

181 182 Table 4.28

Pearson Product Moment Correlations BetweenAttendance and Performanceon the Nuer Literacy Tests

Correlations Between Attendance and TestPerformance

Nonsense Known Group Unknown Syllables Reading Words Words Comprehension Advanced P2s I.M=M11, (N = 15) 00 .16 -.03 -.01 IntermediateP2s

(N= 7) 00 .14 -.15 .56 Intermediate P1s (N 20) -.11 -.72 -.04 .09 Beginning P2s (N = 13) .07 -.12 .11 .34

PupilAttendance In Days

Advanced Intermediate P2s P2s (53)1 P2s (50) P1s (41) (Beginners)(49) .11MIMISMEMMIIIMIMM.1111011.11 Mean 41 34 28 30 Standard Deviation 8.40 10.34 10 8.40 Range 23 52 24- 49 7 - 39 13 -42

1 Numbers in parentheses indicate the total numberof days school had they were tested. been in sessionor each group at the time

183 PAGE 93 consonant in CVVC sequences. They would /lua/for lue k read /gaa/ for gaak (bait). Thisproblem (to argue),or letter-teaching underscores aweakness of the approach, whichemphasizes the SIL pupils areso used teaching of CVunits. to breakingdown wordsin CV units The difficulty soundingout sequences that theyhave very well for like CVVCor CVC.The SIL approach languages withCV/CV/CV syllable may work great many ofthe languages structures;regrettably, be in SouthernSudan are not a the reasonwhy aaore of this type.'This may phonics-orientedapproach apparently yieldedsuperior results to word attackhas Bari comparison on some of themeasures of reading classes reportedin Cziko and for the Cowan (1982,p. 23). Oral reading errorsmade by involve the MuerChildren the substitutionof known words. tended by andlarge to "Unknown Words" With no section, even context, e.g., inthe replalcing a word P2 intermediates that they would steadfastlypersist in word e.g. should have beenable to sound cik "to puta grass skirt out witha familiar the third around a hut," person plural past was replacedwith cike, test passages aspectual marker. followed thepattern J Errors made inreading the graphemic (visual) shape. substituting wordsthat were Sometimesthe recognition similar in graphemes wassufficient to of onlyone or trigger asubstitution, two "fishing", isrendered EJL1 (into) e.g., the finalx in mmy final in (3). graphemesequence Similarly, therecognition of th inbith (spear) in the substitutions in(4b) and Pc), (4a) gave riseto where theth appears the substitutions thal"cook" and thin word initially. Both in instruction "be present,"had been used the pupils hadreceived. very frequently (3) a. Ct wa may yieer he/past go fishing river "He went fishing(in) the river," was read as:

b. Ce WR rey he/past go yieer into river "He went intothe river" (4) a. Man DErn ce Deo Mother of Deng Item bith she/past (name) gave spear "Man Drengave Deo a spear"

b. Man DiE0 ce Deo kem thrtl Mother of Dengshe/past (name)Me cook Man Denggave Dengcook"

and:

grriiitirEWEthe SIL method the fact developed by S. that many ofthe languages Gudshinskyowes something to structure. whereit was John Hollmanalso explained first used hada CV/CV thatanother reason attack strategyisto avoid for adoptinga CV automatically the problem ofthe occurs after the epenthetic vowelthat isolation pronunciation of -- a problem that a voicedconsonant in phonics must dealwith.

185 PAGE 94

c. *Nan Dern cc Den kaa Maher of Deng thin she/past (name) giVebe present "Man Dreggave Den was present" NOre often, the substitutionwas similarity, with based on a closeoverall graphemic therepl2cesentword like the examplesshown Replacements like thesecan be found at all in (5). were made by the advanced levels; (51') and (5h)for example, P2 readers. Notice how close thuuk (finish) anddu the replacement th (remain)are to the printed words words and lu th (get up). in the text, tuk

(5) a. Cike rec DM nag they/pastfish nase take "They took Deng'sfish,"

was read as:

b. Cike rec DO n5n they/past fish (name) bring "They brought Deng'sfish," c. Jok Cc cibat soc ke (nase) luck he/past fish hookgive to frog "Jok baited (his) fish hook witha frog" was rendered as:

d. Joie Cc cibat soc ke (nase) luak he/past fish hookgaveto barn "Jok baited hisfish hook witha barn" e. Nee ci nihiaal ku tuck ke When aspect rain dew begins fall ."When the rainbegan to fall,"

was rendered as:

f. Nee ci nihiaal ku nuuk ke des When aspect rain finish fall "When the rainfinished falling," g. Bi duldu a Loth ke 2nd sing belonging runwagnidiaal... get off in mornings all future asp.your early "You will getup arly every morning. . ." was read as:

h. Bi dugdu a dupth kc 2nd sing runwanni diaal... belonging remain inmornings all future asp. your "Youwill resain every morning. . ."

18 6 PAGE 95 It was often difficult toidentify represent a P1'sattempt to make unambiguously thoseerrors came good semantic thought to to a wordhe could not sense out ofa sentence when language and read. Muer is a he a large numberof the subject-object-verb sentence finally recorded errors (SOV) (6a),or were verbsappearing either reasonablyassuie penultimately (6b). One might, that thesubstitution of forexample, represents anexample of thal (cook) for nen in since the only an effort tomake a (6a) graphemic similarity semanticallyplausible that appeared betweenthe replacement reading, on the test,nag(tike) is word and the both substitutions, the letter word n5k (kilnand thal (cook) a However, inthis case, inthe primer, so it had been used is possiblethat very frequently substitutiononce again, thereader may have on thebasis of made his similarity.The same a remote argusent may be but nonethelessvisual errors like(6d), which made for example(6b). appears to bea bona fide There werefewer substitute a wordthat simply example of the pupil had sade semanticgood sense, an attempt to read. in the contextof what (6) a. Cita rec Deco They/past narj (n8k, thal) fish (of) nese "They take (kill, cook) took/killed/cookedDeng's fish" b. Ciks del nag (to) luaak (nano They/past sheep nak, lath) take barn "They took the (bring, kill,put) sheep to thebarn" c. Cike rig wic *they/past run village "They ran tothe village,"

was read as:

d. Clke rig door They/past run bush "They ran intothe bush" The overall trend revealedby the Muer seen inthe obvious tests was progress made by very encouraging.This is advanced groups the pupilsin Oho hadtheir first th P2 intermediate and 1981. With one exposure to exception, bothgroups scored vernacular instruction in literacy testa.There is thus above 614on all sections and limited every indication of the sources the that despitelarge class the developsent pupils at NasirWest arl making sizes of criticalreading skills. substantial gainsin the effectivenessof the These gainsare a testimony the teaching grouping-by-ability to talents of the procedures usedatthe school, Muer primers. IRL LanguageOfficer and last, but notleast, the

In summary,the Zande and picture than &ter outcomedata present that reportedfor a far more second the Bari andLotuho pupils encouraging impact evaluations. In terms of in thefirst and Zande and Muerpupils appear overallliteracy and Lotuho to enjoya definite achievement, the counterparts andthis must in superiorityover their Bari in the Cycle2 materials, large part bedue to the in the which have donemuch to promote innovations classroom.Clearly these Jeaningful reading Cycle 1 materials innovationsshould guide as well as thecreation of the revisionof the new primermaterials.

187 PAGE 96

4.4DISCUSSION

As is evident from the preceding text andtables, a very large data were collectedduring amount of the threeimpactevaluationsof the project.While the preceding literacy sections of this Chapter haveserved primarly to present these data, thepurpose of the present section is evaluation findings. to discuss the This is done by makingsome general comments on found out about primary what was edncatIon andliteracy inthe ruralareas of Southern Sudan. Thisis followed the by discussions ofthe overall literacy achievement of primary schoolchildren in the region, evident of the impactof the literacy project andthe implications of these findings for improvingthe literacy achievement ofChildren in thisand similar third world. regions of the rural

4.4.1 PRIMARY EVALUATION AND LITERACY IN THE RURAL SOUTHERNSUDAN

It is apparent that there are many serious obstacles tothe acquisition of literacy skills by primaryschool pupils in first set of these the rural Southern Sudan. The obstacles arerelated to school factors. prolongedteacher strikes, Frequent and untrained teachers, highteacherandstudent absenteeism, lack of bothteacher and pupil materials, and occasionallyvery large classes and poorfacilities all impede primary education in thisregion and literacy acquisitionin particular.

In addition, the methodsemployed to teach reading and writingprincipally involvethe mechanicalrepetition and memorization Absolutely no "meaningful" of words and stories. activities took placein the vernacular classesof the two comparisongroups. Questions requiring readingcomprehension were not asked, themeanings of stories read were not discussed and the functional aspects of literacy (e.g., reading and writing correspondence, instructions, record following keeping) were notdemonstrated. Second, in areas visited there appears the rural to be virtually no contactwith outsid, of the context of the printed word the school and no use of literacy skills by other %Ianthe teachers, government anyone officials and police. Furthermore, any contact with literacy is likelyto be in Arabic orEnglish and not in the vernacular language. This is in stark contrast to most urbanand developed areas of thewcrld where childrenare often as immersed in thelx language as they printed forms of are in spoken language andsee reading and writing daily the functions of as an integral and meaningfulpart of their environment. Since these factors willmake it extremely difficult forany literacy project to have an impact in the area, this project can be viewed as a testcase for literacy. Totheextent that these conditions aregrinner toother economic...211y undeveloped areas of the world, the results of thisproject provide anindication ofthe impact that a vernacular literacyproject is likely to have inrural, underdeveloped areas of the worldwhere there is no or very limited contact withor use of literacy skills.

188 PAGE 97

4.4.2 OVERALL LITERACY ACHIEVEMENT

The overall literacy achievement appeared to be generally verylow with relatively few exceptions for allpupilstested over the three-year period of the project evaluation. Among the P2 Lotuho comparison pupiletested as part of the first impact evaluation; only one child was able todemonstrate any literacy skills at all (he was able to read aloud one word from List1 and one word from the story passage).

While the test results presented in Table 4.7 do indicate thegenerally low level of literacy achievement, the literacy skills of these childrenare best appreciated by listening to the recordings of their oral reading ofthe story presented to them. Eventhe Bari and Lotuho pupils whowere most successful in reading thestory read painfullyslowly withlittleor no indication whatsoever thatthey were understanding whatthey were reading or were able that they to use syntactic or semanticconstraints inreading. while the Therefore, Priaary 1 pupils of the Bari,project, Bari comparison and Lotuho projectclasses had acquired some wordrecognition and basic wordattack skills (the latter best demonstrated by the Bari control group), theirreading comprehension skills appeared to bevery poor indeed and far below the general level attained.in the US by childrenat the end of Grade I. The follow-up testing of theBari and LotUho pupils at Primary3 in the second impact evaluation generallyconfirmed the results of the previous In addition, year. the detailed analysis of theoral reading errors of the Bari pupils revealed that at the end of their third year of school theywere usitg almost exclusive visualinformationin reading(typical ofvhildren just beginningto read) and wereless sensitive to the useofcontextual information in reading than typicalAmericanchildren toward the end of their first year of schooling.

The performance of Primary 1 and 2 Muer and Zandechildren as assessed as partof thesecondand third impactevaluations was considerablymore encouraging. Whilethe PI Nuer projectchildren atNasir included in the second impact evaluation appeared to havequite poor reading skills, the P2 children tested at Noir the following year faredconsiderably better. More lapressive still was the performance of the Zande project pupils whoappeared to have developed the highest level of readingof any of the groups testedas part of the three impact evaluations.

In summary, while we must becautious in making comparisonsacross language groupsdue topossible differences inthe difficulty of MAT tests, it does appear .14ospective that there was a distinct difference in ?sew .ng skills between the Cycle 1 (Bari and Lotuho)and Cycle 2(Nuer and ALI') groups which were evaluated with theformerchildren ata very L4 levelof achievement and the latter groups showing evidenceof having tttainedsome basic reading skills, as shownby the abilityto read unfamiliarwords and devise some comparison from the readingof simple passages.

189 PAGE 98

4.4.3 EVIDENCE OFIMPACT OF THE PROJECT

One of the primary reasons forundertaking the evaluation chapter was described in the to assess the impact that the project was having developaent of primary school on the pupils' reading skillsin the Sudan. However, rural Southern in spite of the large amountof data collected three impact evaluations it during the is not possible to stateunequivocably whether project at the time had the or did not have apositive impact on achievementof the the literacy project Bari, Lotuho, Nuer and Zande pupils. obviously thecase for Thisis the Nuerpupils sinceit wasnot possible non-project Nuer pupils who to test were not using the new projectvernacular literacy aaterials. Even forthe remaining 'three language groups wherecomparison groups were available for testing,the lack of any type of control of strict experimental the groups (Mach as randomassignment to project classes) makes impact and non-project assessment risky sincedifferencesbetween project and comparison groups may be due primarily to theother contounding uncontrolled variables and not to the projectitself.

Nevertheless, it is informative to compare theperformanceof the project andcomparison pupils in readingBari, Lotuho and Zande cooper/2,ns, along with sincesuch consideration of the background each participating data collected for class, do provide theonly objective impact of the project evaluation ofthe on the development of literacyskills.

The project and comparisondata collected from the thefirst and second Lotuho groups as part of impact evaluationsdo suggest that responsible for the the projectwas generally better performanceofthe Lotuho project pupils compared P2 andP3 to the.non-project classesevaluated and 4.13). (see Tables 4.7 However, it must be recalledthat the Lotuho comparison not ableto offer its pupils school was individual Lotuho languagebookswhich limited all reading instruction towhat was writtenon the Chalkboard. assume that most Lotuho However, if we primary schools in Southern Sudanare Lotuho comparison school similar to the evaluated, but have no or inadequate pupil materials for teachingliteracy, then the differencebetween the Lotuho projectand comparison groupsin reading performance could be taken as anindication of the impact of the literacyproject.

However, the relatively goodperforaance ofthe Bari comparison Tables 4.7 and 4.12) suggests group (see that whatmade a difference is theavailability and use of aaterials by thepupils.There was no clear evidence project materials that the Bari being produced were superior to the olderaaterials being used by the Baricomparison group. In fact, the teaching ofletter names to the Bari comparisonpupils seemed to have given this group an edgeover the Bari project group in soundingout and recognizingnew words.

The final source of dataconcerning the impact of the the comparison of the project is given by Zande project and comparisonpupils provided third impact by the evaluation (see Tables 4.24 and4.25). Here therewere quite clear differences in reading test performance infour of the project pupils. There were also differences in the pattern of oralreading errors suggesting thatthe project pupils had developedbetterskillsin using syntactic and semantic contextual constraints in reading thandid the comparison pupils whorelied on primarily visual information, a characteristicof beginning

190 PAGE 99

readers. Again, however, it should berecalled thatthe Zandecomparison pupilsdid nothave their ownbooksforuse inlearningto read language. their

The Lotuho and Zande situations are roughly parallel in that in bothcases the comparison schools did not have adequate materials forteaching literacy. Hence it can be maintained with* reasonable amount of confidence that these circumstances under the literacy project hasdefinitely had a very positive impact. It -is not clear that theproject will domuch toimprove primary school literacy in areas where competing materialsare available trained teachers. and used by However, since this situationappears rare in rural areas, it isobvious that continued implementation or theproject will domuch to increase primary school literacyin the Southern Sudan.

Thus, two general conclusions can be maderegarding theimpact ofthe project. The first is that the project hashad limited butnonetheless positive impacton the development of vernacular literacy skills for those children who would have otherwise not have had accessto vernacular language reading materials. A second, more cautiousconclusion, based on what appeared to be a higher level of reading skills among the Muerand Zande project pupils compared to the Hari and Lotuho groups, is that the modificationsmade to the Cycle 2 materials appear to have aadethese materials more effectivethan the Cycle 1 aaterials in fostering the development of readingskills.

4.4.4 IMPLICATIONS

The findings of the three impact evaluations lead to numerousimplications concerningimportant considerations in attempting to improvethe level of literacy of primary school Children livingin rural, undevelopedareas of the world. Here we will consider the school-related factorswhich appear to influence the learning of literacy skills in schools located in suchareas and recommendations are offered to increase the effectivenessofthe teaching of literacy skills in this and similarprojects.

The two major school factors which have an obvious impact on the teaching of literacy skills are the teachingtechniques used and the readingmaterials available tothe child. First with respectto teaching techniques, it was found that almost all classroom activities, both in the projectand comparison schools, involved the repetition of . syllables, words and sentences chantedby allChildren inunison as they read fromthe chalkboard. This isnot surprising when it is considered that few teachers had anyformal teacher training experience and that few of the comparison schools couldsupply their pupils with reading aaterials. However, evensome ofthe teachers of the project classes which did have a generally adequate supply ofpupil primers relied heavily on copying reading lessonson the chalkboard (which often took a good deal of time) and extensive unison repetitionofthe reading lessons. None of the activities observed in any of the comparison classrooms andfew activitiesobserved in the project classrooms involvedreading for comprehension. As noted earlier, reading comprehensionquestions of the new project materials could usually be answered bya repetition of thatpart of the passage relevant to the question. It is therefore understandablethat all

191 PAGE 100

thecomparisonpupils and most of the project pupilshad considerable difficulty in answering the reading comprehension questionsincluded evaluations. in the

The project asterials have been describedin detailin thepreceding chapter whereit was feared that the extremely hip rate ofrepetition of wordswithin theprimers aa well as the limited umber ofinstructional activities requiring readingcompreAension would likift lead to building a sight some skill in vocabulary ofthe oft-repeated saris batpoor skills recognizing new words and in reading in comprehension 21a.The performance of the project pupils on the evaluations' reading testa Zs* generallyconfirmed this suspicion. In addition, the changes made to the title 2 Muer and Zande materials in an attempt to provide more meaningful reaftsexercises was found to provide the pupils withmoremotivating and stimulaWing reading which may be at least partly activities responsiblefor the migarentlybetter reading skills of the project Nuerand Zande pupils compursi 'to Lotuho children. the project Bari and

Finally, there is a clearneed to make verracularreading skills useful both within and outside school if these skillzare to be maintained by children. well developed and Within school,vernasalmr language materials teaching of other subject for the areas would prowl& a strongmotivation development and use of vernacular for the language rszellng skills. Outside of school, attempts need to3 be made to involve parents andother adultsin vernacular literacy projects and to demonstratethatliteracy can serve a useful in village purpose life. Indeed, the generallybetter performance of the children in the larger project towns of Yambio and Toritmay be due to the fact that these Were the only two coamunities studied where the written wordcan be seen many places outside of school.

In light of these observations, the following recommendations appropriate. seem While these recommendations havebeen made with this particular literacy project in mind, theymay well have value in other undeveloped settings where attempts rural are made to teach vernacular literacyskills. 1. Considerably sore time should bespent on meaningful readingactivities inthe classroom. These activities could include the raading and comprehension of more interesting and variedstoriesrelevant tothe experiences of children living in rural villages, discussion ofthese stories, allowing pupils to create, write down andread stories of theirown and the writing of letters to friends and relatives in distant villages. The apparent success of the context clue drill included inthe Muer andZande materials also indicatethat more meaningful activities can become part ofthe highly structured foraat of the vernacularprimers.

2. Questions used toassess reading comprehension should such a way that they could be constructed in not be correctly answered byverbatim repetition of parts of the story or by generalknowledge of the local culture. stories in Many of the the project aaterialsinclude three major characters:a woman and an animal. man, a Therefore, questionssuch as who was killed (inevitably the animal), who killed it (inevitablytheman) and who cooked it the woman) do little (inevitably more than test memory for the names of theman and woman (oftenthe samefrom story to story)and the type of animal involved.

192 PAGE 101

Comprehension questions should involve inferencingskills and elicit discussion of the meaning ofthe story.

3. Better use shouldbe made of class time. Relevant to this recommendation is the findingthat the two project time writing materials teachers spent considerable on the blackboard whichwere already pupils' materials. contained in the This wasted time shouldbe used activities. in meaningfulreading

.4. Materials need to be producedin the vernacularlanguages for the teaching of other.subjects. The learning of other the vernacular subjects through readingin would demonstrate tothe pupils that reading is a means for obtaining information andnot simply a mechanical own sake. process performedfor its

5. The only reasonablesolution to the problem of large classes isto break them up into millergroups. Where there is a lack oftrained teacheis andan abundance of pupils, the approach taken at Masirrecruiting and training teachers' aides --seem tobebetter than letting students. one teacher handle40 to 100 Some attention shouldbe devotedto developing a procedure training teachers' aides. for

6. The comparison of the Zande and Muerchildren's reading abilities constitutes a compellingargument for lAstituting children are allowed to a policy whereby the retain their primers outsideof school.The potential benefits from such policy far outweigh objections suchas possible expenses incurred due to lostor damaged books. Although it seems clear thatallowing pupils to bring their bookshome will not in itself of fluent reading skills guarantee the development (since the Bari andLotuho project allowed to take their pupils were also books out of school), itis the only way which these pupils can have most of contact with literacyoutside of school and the by which their parents and other only way faaily memberscan become involved in and stimulate the education ofthese children.

7. Finally, in settings suchas these studied herewhere littleif any reading materialsare available outside of school and where few peoplecan read, efforts must be madeto provide sources of vernacularreading materials and to allow adults,particularly parents, to become literateintheir own language. It 99023 clear that children will learnto read 'best in those communities which haveat leastthe beginnings of Further a climate of literacy. thoughts on how literacycan be made to meet the undeveloped communities will needs of rural be offered as part ofthe next chapter.

193 PAGE 102

5.0 GENERAL DISCUSSION

The purpose of this chapteris tosummarize the major lessons that learned from studying the may be development of the Southern SudaneseLocal Languages Literacy Project by relatingthese to the broader contexts oflarge-scale literacy programs and languagepolicy and language planning.

The results of the impact avaluationssuggest that thesuccess or failure of large-scale literacy projects in developingcountries is in part dependent upon variables which have been traditionally identifiedas critical components of all formal educationalendeavors, i.e., teaching amount and type of training aaterials, methods and the the teachers have received.The apparent improved performance ofthe Cycle 2 pupils over the Cycle 1 pupils ismost likely attributable to the difference in the quality of materialsused bythe two groups, since potentially confounding variables appear to be fairly equivalent.This argues that materials can make a difference in thelevel of proficiency attained even whenthe setting of a literacy areas of a pre-literate society. projectis in rural Materialsmost conducive to promoting functional literacy. would employ the technology of linguistics for the creation of optimal orthographies but would alsoincorporate pedagogical applications of insights derivedfrom up-to-date reading importance of episodic structure research, such as the for comprehension andretention of larger prose passages and the necessityfor developing in the how different kinds of learner an awareness of contextualinformationcan assist in unfamiliar words. deciphering

In the classroom, activities which promotereading for meaning should precedence over take teaching techniques whichoccupylarge numbers of pupils in mechaniaal operations,e.g., the sustained repetition ofsyllables, words, sentences, even when suchtechniques have an established culture. tradition in the This priority of emphasizingthat reading is for meaning just mouthing sounds which rather than have symbolicrepresentationon a page blackboard should carry or over into teachertraining courses, where, becauseof the trainees' often meager forsal education, thereis a tendency inmany African countries to equipthem with a limited number of which focus almost exclusively teaching techniques on establishing sound-symbolcorrespondences (five teaching techniques wasthe maximumnumber setby the SIL training expert). teacher It is obvious thatoneshould not attempt to load in this setting with teachers a complex set of techniquesmore suitable to developed countries, and, similarly, itis clear that all beginning readers in must engage aome word-attack practiceif they areeverto master the principle, but alphabetic this shouldnot exclude the possibilitythat at classroom least some activities emphasizethat readingIsfor meaning. The lack of classroom techniques emphasizingreading for meaning was undoubtedly weakness of the Sudanese project. a serious

The total number of external variables which deserve serious when attemptingto mount a successful consideration large-scale literacyproject will unquestionably vary from country to country, althoughelementscommon to all similar settings may well be found in the Sudan experience. It is clear that

194 PAGE 103

the project has produced much knowledge which may be appropriatelytransferred tofutureendeavors in preliterate ruralsocieties. Thestatistically significant superiority of the Pi Zande pupils who were allowedto take their primers home with thee after schoolover those pupils who were not one of the most argues that important variablesin fostering reading ina preliterate society isthe creation of a climateforliteracy inthe Governments engaged in community. multilingual programs similar tothe one described in this report should realize thatimplementation of rural literacy sutztantial long-term requires a committment at two levels. Not only must vernacular materials be made available in a wide range of schoolsubjects, i.e., get:graphy, history, math, etc.,but some inducement to readon a daily basis must be created within rural communities ifthe literacy training going the schoolsIs to have any on in relevance in theeyes ofthe childrenwho are receiving It.The total investment involved in the successfulimplementation of rural literacy caapaignsin preliteratesocieties is required thus doublethat in countries where thevehicles for promoting and literacy, sustaining i.e., newspapers, journals,books, signs, etc., are a feature of the regular community environment. Government officials literacy programs contemplating in countries with conditionssimilarto thatof Southern Sudan should not lose sight ofthis very obvious but fundamental point, it implies that (a) since a greater financial allocationmay benecessary than was initially anticipated,and (b) the cooperation of sore than agency, one government e.g., theMinistry of Education and theMinistry of Community Development will need to be solicited, something whichis notaccomplished with ease in the most efficientgovernments.

Nevertheless, the results of the SudanProject clearly point to the need for establishing literacy inthe coamunity prior to or in tandem with primary school instruction. One of the most importantreasons solittle learning is sustained from year to year inSouthern Sudanese rural schools the lack of reinforcement appeared to be available in the community andthe consequent poor school attendance of the children. This was not due to a complete effort in this area by the lack of IRL/SIL personnel. One of thegoals of the Local Languages Literacy Projact was theinstitution of writers' workshops regional centers. in These were held periodicallyby SIL literacy consultant Wanda Pace, their purpose being tostimulatelocal writers toproduce indigenous literature of any sort --usually folk tales. Unfortunately the inability of the Southern Regional Ministry of Education to provideresources for the production and distribution of the materials whichevolved from these workshops nullified this promisingeffort.

However, there is a differencebetween creating an interest in ensuring its acquisition. literacy and Recall thatin Chapter2 it wasdiscovered that although 94% of the people surveyed felt rsadin, end writing to bedesirable, rise major preference expressed wasfor litel 'e which had noinzediate r.iactical benefit, i.e., tribal histories and f(s.. lore. The history of the evolution of literacy (Gelb, 1956; Schmandt-Besserat, 1978) showsthat reading andwriting can only be sustained withina societyif theyfulfill an importantpractical function. Thus while the writers'workshopswere undoubtedly a step in the right direction, they wouldhaveto be followed by transitional efforts designed toispart therealizationthat reading and writing havesomeismediate useful purpose in theseruralcommunities. Typically, the next stage in suchprograms is the introduction of literature ontopicsdeemedbeneficial for thecommunity as a whole, e.g., the

195 PAGE 104

improvement of hygiene, agricultural yield, etc. But, as was Dr. Singh, the UNESCO Community pointed out by Development Specialist sent to Nardito set up the first of a series ofIntegrated Regional Educational often requires a fairly long Centers (IRECs), it passage of time inpreliterate societies before the written word becomes an accepted mode of transmittingknowledge. (This was the reason why literacywas notcontemplated as a means of the IRECs.) As a result, some instruction in in-depth preliminaryinvestigation (perhaps using a trained anthropologistas was done in this study) should be to determine ways in which undertaken literacy can be consistentlyuseful and relevant in thelivesof rural preliteratesocietiesand the findings investigations should of :11ese then shape effortsby the government to reading and writing within establish communities on a more permanentbasis. Certainly one of the major lessonsto be learned frda theSudanese experience isthat theimpact of multilingual literacy programsmounted in preliterate societies isdependentupon the interaction of sociological, economic and political variables.Sporadic attendance by to be due not only to the teachers was found fact thatthe profession is poorlypaid, alaost universally, as it is but also to aconflict between the demands teachers by a western educational imposed upon system endtheir desire to maintain traditional economic patterns. Similarly, pupil truancyarose .from a number of factors: the need to utilize children from time to time asa member of the family work force, a conflict between the traditionalsubsistence schedule and the western educationalschedule, the lack of understanding what by parents as to constitutes consecutiveattendance andit's importance and apprehension about parental the possible use of the classroom as a vehiclefor political, cultural and religiousindoctrination. Dealing with these conflicts effectively willnot be tasy. Some maybe alleviated by increased of them financial commaitmentto public education, e.g., in Sudan,raisingteachers salaries mightattract personnel and reinstituting more qualified boarding schools at theprimary level would do much to guarantee more consistent attendance.But more aoney is not the answer.Other only conflicts-that arise from the three factorsmentioned above require sophisticated approaches,since theyinvolve changing behavioral patterns that have guided people's lives for centuries. This the knowledge which perhaps takes tine and only acultural anthropologist witha sharpened historical perspective can supply. But it should be clear that done to enlist the nothing can be community's increased support offormaleducation until theirperceptions and apprehensions about schools have beenidentified. Educational authorities musthave a fullerunderstanding of the rural population's perceptions of the school'sfunction, why parents send children to their school, how large a say theywant in what goes on in what their committment to class and the school is in terms ofreleasing what for them is asignificant workforce resource. Goverments should understand acquiring trustworthy that answers to these questionsis important since suchdata can bearsignificantlyon the effectivenessof strategiesthey maybe contemplating, and, hence, constituteimportantindicators of aliteracy project's potential for success. The most sensible could save governments course of action, which considerabletime and money in the long run, would be to collect dataon the sociological, economicand politicalvariables that bear on theproject as a first step. This would permit Oolicy planners and educational authorities charged with mounting a given literacy project to

196 PAGE 105

recognize some of theotherwise more formidableobstacles they faceand to consider how they may be mosteffectively dealt with.

One of the more reliableconclusions to be drawn is that the evaluation from Sudanese experience of large-scale literacy projects being implemented in rural settings generatesmore questions than unequirmalanswers. Early in this study the need forbroadening thedata base by including more schools became evident. children from The difficulties involvedin developing and administering objective tests inruralAfrica meant comparisons of achievement thatstatistical tcoresby differentgroups interpretod for descriptive couldbe best purposes only (dueto small sample sizes, of trulyrandom selection of exaninees absence and thefact that schoolsare often chosen primarily because oftheir accessibility and hence could not alwaysbe considered representative). Inthecourse of numerous interpreting the testdata questions were raisedthat appeared tobe expanding the answerableonlyhy data base and collectinginformation about the hidden sources of education possibility of in the community andparental support schools. It was at this point for the that the SIL anthropologistRobert brought in. Much of the Hoppe was intonation he collectedwas already known to the consultants andin fact had been touched on briefly in some of impact evaluation reports. the earlier However, some of theobservations produced by his survey, particularlythose pertainingtoparents' conception of what constitutes adequate attendance by theirchildren and their function of the school views ofthe in rural society addedgreatly to of factors affecting the our overall picture total lapact of the literacyprogram.

The experience gainedfroa this particular project suggeststhat the most appropriate evaluation designfor similar large-scale literacY where programs is one variation ofthe previously mentionedvariables is variation in used toexplain the attainment of literacyskills and school achievement. evaluationof such programs would The thuscombine informationderived quantitative instruments and from qualitative impressions. Needless tosay, the naturalistic variationdesign advocated here requires a long time frame and evaluators with complementaryskills.

Ideally, one would want toemploy the talents of a sociolinguistbefore embarkingon any action. The sociolingu!st would assess thedegreeof bilingualism in the areas wherethe project i3 to be mounted, theplanners to make thereby enabling justifiably well informed judgments as languages should be targeted to how many for literacy development.The initial SILsurvey had attempted to do thisand had brought into focus nine Role A for priority attentionthe languages. These languages wereidentified as suitable for literacy development the ones most and as the languages whichwould cover the aajority of the childrencomsencing school in the South.

At the sametime an anthropologist could collect dataon the sociological variables which interface with the introduction of adultand primary education in ruralareas.Armed with this information theplanners could sapout an appropriate course of actionfor implementing the project.Applied linguists with a specialization in reading could thenbeemployed to developing materials and, assist in finally, an evaluationspecialist the anthropologist tO could work with ensure that useful foraativeevaluation the materials developers is funneled to and teacher trainers.While the scheme proposed here may well sound far too elaborate tobe taken seriously byproject development PAGE 106

experts, itshould be emphasized that alternatives,e.g., just using one or two specialists, runs therisk of presentingthe projectdevelopment staff with a very narrow viewpoint thatmay well fail to take in important variables which should be addressed.

In reviewing the Sudanese literacy project one inevitably feelscompelled to address questions such as: Are multilingual literacy projects ofthis type justified? What does theoutcowe of the SouthernSudanese, project reveal aboutthe potential formultilingual educationalpoliciesin thirdworld (particularly African)countries? The answers to thesequestions are obviously to a large extent, matters of individual interpretation basedon the interpreter's point of view and henceare not neatly resolved by citing recent or past evidence. In attempting toanswerthe first question one can confine one's argument to the Sudanese context, as Mahmud (1983) has done,or adopt a more global, historical perspective as is seen in the reasoning employed by Bokamba (1983, 1984). In an admittedly political treatise(which nevertheless lays claim to historical and empiricallyderived support), Mahmud contends that the spread of Arabic as a lingua franca, destinedeventually to replace the many individual languages in theSOuth, is inevitable.Using data from a survey of 2894 students enrolled ins21 schools in the capital of theSouthern Region, Juba (aad one school outside of Juba) Mahmud states thatthe first stage of an irrevocable process, fostered by marriage among different tribes, whereby Arabicbecomes used withincreasing frequency to meet communication demands has already begun. The obvious implication of thisthesis is that effOrts.of the IRL are a waste of timesince they only tend todelay the inevitable domination of Arabic. The factthat datacollected from the :argest city in the South might notbe generalizable torural areas seems to bother Mahmud onlyslightly, since he mentions itfirst as a potential limitation of his data, but then dismissesit subsequently claiming "vhatever differences exist among sectors are essentially ones of degree andnot of kind"(Mahmud, 1983:61). The invalidity of thisunscientific approach to theory formulation is borne out by Hoppe'sdata, which shows that only in the most northern province, Bahrel Ghazal, isArabic making any headwayas a preferred language.

Whether Mahmud's prophecy will be fulfilledeventually is an open question. It is probably more likely thatArabic willassume a wider roleas a lingua franca than it presently does, withoutreplacing local languages completely. One can point to counterexamples in Sudan today where tribesin the Northern Region have resisted acculturation and assimilation and kept their language83 theprimary means of communication. The mostnotable example of this is perhaps the Gaam, a tribe of about 40,000inhabiting the central.massif of Blue Nile Province, who are often referred to as the "Ingessana", presumably "the thankless ones" for their refusalto accept Islaa. Still, the spread of primary education to rural areas offersthe mostlikelyvehicle for the expanded use of Arabic, ao Mahmud'a thesiscannot be rejected entirely.

A conceivable long-term resultof the propagationof Arabic through the educational system is the extinction ofminority languages presently spoken in Southern Sudan. Although this possibilitypresently appears somewhat remote, it is not without historical precedent. The maintenance of anylanguage depends on the extent to which it is actively used by successivegenerations. Whether future generations continue to acquire a language isdetermined by a number of sociolinguistic conditions in the area where the language isspoken. PAGE 107

Among these are the percieved status of the is language, the.extent to which it spoken at homeand in theeconomic community, andwhether ithas socioeconomic value, i.e., isused to transact business. Languages perceived as having no valueother than the maintenance likely be abandoned of a cultural heritagewill by future generationsfor currency brings greater languages whosefunctional rewards. Dressler (1982) points speakers begin outthat once to perceive theirlanguageas having diminished value, this perception prestige and is difficult toreverse. There of languages which have are numerous examples been subject tothis kind of prestigeloss resulting diminisheduse: with a Irish in Ireland (Macnamara,1970), Welsh in Wales (Lewis 1982),and Yiddishin Central and EasternEurope (Fishman, 1982). Similaeoly,there are severalcases of language decay and traced to death whchcan be the elevation ofa single languageto the status of language, e.g. the a national decline of Greek andLatin in the centuries (Kahane & Kahane, fourth andfifth 1979), the decreaseduse of Slovenian in Corinthia and Breton in France (Brosnahan, 1963), and the extinction of hundredsof Amerindian languages in NorthAmerica (Leap, 1981). The possibility that a similarfatemight await hundreds of languages isrecognized by Bokamba who African both as a geographical argues that "if Africa is to develop and cultural regionit must heritage. And preserve its cultural thisculturalheritage is in its (therefore) be preserved and languages...whichmust enriched, because theyare the tools of personal and national most effective development" (1984,p. 39). that although very few linguists He goes on to note and anthropologistsencourage thedeath of languages, . . . "'the multiplicity of Africanlanguages makes anyrealist welcome such a developmentfor national integration this context purposes." It is within that Mahaud's polemic Isbest understood. hiaself one of the Heundoubtedly sees "realists" whofeel compelled to seek viewpoint that might well be support fora accepted as "intuitively"correct by many African governments, i.e., that multilingual literacyprograms are a luxury which contries seeking to promote national integration cannot afford. Rather then expend resources on suchprojects, the official the first year of school. language should be taughtfrom

The superiority ofthisunilingual approach,which has acceptance among Sudanese gainedsome educators (see for exampleScharrer, multilingual approach, which 1983), overa makes use of the child'smother tongue in school, has not been demonstrated. There are a number of children can learz to read studies which show that a second language whilelearning to read their (Cohen, et al 1970; Bowen, own 1968; Lambert andTucker, 1972; Malherbe, 1946; Ramos et al, 1967), and thuscalling intoquestion thevalidity of pronouncementthat a child should first the learn to read andwrite in the language spoken at home if heis tocommand any language (UNESCO, However, in support of 1953). the multilingual approachit should benoted that the excellent results obtained inthe aforementioned part to sociological studies may be attributedf.n status (the childreninvolved came from aiddle classhomes) and upper middle and in partto thequality or the programs excellent facilities, (they had materials and highly trained teachers). where these conditions are In countries not present, controlledstudies have shown that beginning reading inthe mother tonguebefore switchingto the (second) language national produces higher levels ofcomprehension instruction in the latter than starting from PI, e.g., Berrera-Vasqueh(1953), Burns (1968), Mondiano (1966), Rubin (1968)and Valencia (1970a, 1970b).

199 PAGE 108

Further support for the multilingual approach can be foundin recent high rates of literacy claimedfor the campaigns aounted in Ethiopia (Demos,1984), which cite theuse of mothertongue literacy as an important factor in bridgingthe gap to learniagtoread the nationallanguage, Amharic. Similarly, one can pointto the success of multilingual programs inthe Soviet Union where various languages of the Republics are usedfor instruction with Russian being taughtas an obligatory subject. believed, have been quite The results, if they can be phenomenalalmost 100%literacy within halfa century. Finallythere is onerather dramaticpiece of evidencewhich suggests thatbeginning primary schoolwith the national languagemay bring aboutundesirable results. In a detailed policies of Vest Africa, examination of theeducational Bokamba (1984) has shownthat the approach Francophone Africa whereFrench is emphasized pursued in to the exclusion ofthe regional languages haa resulted inan alarming increase in decades. illiteracy over thepast two

Although as we haveseen, there is evidence that both monolingualand multilingual programscan result inthe attainment language, it aay be that of literacy in second the fo:mer aremore effective in highly societies where there is developed parental pressure to supply motivation whereas thelatter are more successful for learning, inless developedsocieties. appear to be definitesimilarities in There terms of setting andworking conditions between the Sudaneseproject and other rural educational programs which have achieved a high rate ofliteracy by beginning reading in thenative language, most notably Mondiano'sChiapas Mexico project, TarascanIndians and Berrera-Vasquez's studyof Burn's project withQuechuan Indians. recent progress shownby the multilingual This and the used to argue that the approach tried in Ethiopiacould be Southern RegionalMinistry has quite the educational policywhich will prove possibly chosen optimally effective forthat area. Advocates of multilingual education alsoclaim motivational advantages that thereare .hidden to beginningliteracy in (1873) and Spolsky (1972), the mother tongue. Preck have observed thatmother tongue minority languagesboosts their instruction in status sinceit implies the acceptance their cultures. The result, they claim, of is an increase inmotivation for the children of these minoritylanguage groups to visits to become literate.On one of his the Sudan, Cowan foundsome anecdotal evidence first part of thisproposition. in support of the While visiting Maridihe had the opportunity to observe a receptionby the tribal elders of assigned to work the S1L linguistwho had been onBaka, a smaller RoleB language which larger Role A tribe, the is surrounded bya Uncle.Although most of the Largely to details regarding discussion was confined how this personwas tobe housed, repeated references to there were the importance ofhis arrival. He was viewed almost a liberatoror redeemer.The IRL as language officer shepherdingthe linguist through the ceremony,aBaka, explained the said: reason for this attitudewhen he "Now we shall havea written language like the longer consider us inferior." Zande, and they can no It is difficult tobelieve that the the Baka will notreceive considerable children of learn encouragementfrom theirparents to to readand write whenaaterials are eventually language. produced in their

The following conservative conclusions may be drawnfrom the discussion.Althoughthere is no definitive foregoing proof that learningto read in the native language firstsakes it easier to learn to reada second language,

200 PAGE 109

there may be benefitsin theform of increased motivation justify adopting to learn which this approach inliteracy programs designed like those for condition4 in ruralSouthern Sudan. An alternative using approach, thatof Lot indigenous languagesbutinstead beginning instruction language atPI, may not necessarily in national prove moreefficient since itwill not motivate children to learnto the same degreeas the these approaches former scheme. Both of will be equally impeded byfactors existing countries such as in developing poorly trained teachers,poormaterials and socio-economic conflicts that interrupt thecontinuity of education. frame--at A reasonably longtime least a decade--will te requiredto evaluate aultilingual approach has whetherthe a reasonable chance ofsucceeding. funding for this project Termination of by USAID deprived thatagency of a unique test which couldhave been held up ease as a standardby which the value of fUture endeavors of this sort couldbe judged.

There can be little doubt ths0; the Southern Sudanese literacy project has been viewedquite differently by the central government inKhartoumthan by the Southern RegionalGovernment, and that considerable attached toit by both parties significance has been ovorthe past decade. The reason why this project is seenas having serious consequences for thepromotion of national unity may best be understoodby viewing it within of language the context of the framework policy decisions proposed byFishaan (1971). factors affecting the He noted that the choice of language policymade could be conveniently divided by developing countries into three types.Type A decisionsare dictated by a generalconsensus among the "leadingcircles"or "elite" thatno "indigenous Great Tradition"existed which the majority of thepopulation might draw upon to assistinpromoting national Fishman cohesion. Insuch cases, suggested thelanguagepolicy chosen bya nation language of wider coamunication is generallya or international language,which is frequently the language ofaformer colonial master. Type B decisions,onthe other hand, are, based on therecognition of the which is existence of a "Great Tradition" "available toprovide the indigenized and laws, symbolically elaborated customs, literature...identity appropriate for nationwide identification."Thisperception leads to the indigenized choice of an indigenous or language. Type C decisionsare characterized refers to as a "conflicting by what Fishman orcompeting multiplicityof great traditions," noneof which dominate. Since each of the traditions is "numerically, economically andideologically strong enough tosupport large-scale sociocultural andpolitical-operational integration" a nation can result in their co-occurencewithin considerable tension. To reduce this, a compromise choice is made wherebya language of wider communication is typically adopted as thenationallanguage andmultilingualism intheform of selected indigenous languagesserves the regional traditions.

Quite anumber of Africannations can be viewed as having madelanguage policy decisions whichfit Fishman's typology. Many ofthe Francophone West African countriescould be considered Type and Ethiopia A nations, andTanzania, Somalia could be classifiedasType B decision countries, Bokamba (1984) has although, as pointed out, Tanzania andEthiopia might be viewed C nationsfrom acultural and a linguistic as Type perspective, sinceother major languages and cultureshave been subsumed under the presentlanguage policies of thesetwo countries, each of which has favored a locallanguage, i.e., Swahili inTanzania and Aaharic in Ethiopia. Nigeria is areasonably good example of a Type Ccountry.The language of Fider communication, English, is

201 004r Atli!4 royffripd Iia

itII

PIgride,11;1118111'hriu!iirelliclitim;;I4hiI PAGE 111 4.0 NM=

Retailed pins are seldom implementedin Southern Sudan without ensiderable mediftlation. This has certainlybeen theexperience of the legal lmegmages literacy project.The early plansproved to be much too eptisistie, but without optimism thereis nohope for the education system isethsre Was. in

Op te Illy 1103, the projectna still progressing basically according the er to salplans, albeit at a much slowerpace than originally expected. the OPIfirel timetable vas revised and slower pace was accepted. This appears be the situation in the great aajorityof projects in the region. Progress enerdimg to schedule presupposes a state of development notyet resoled la Sudan.

Primers hod been written in seven ofthe Role A languages and teachershad been trained to use them. An accurate record of schools and pupils using bens IS 110, the possible, but the numberWas increasingsteadily andmore tesehera were trained In the early monthsof 1983 for the new schoolyear.

Per the too remaining Bole Alanguages, an SIL linguist wascarrying out research to eme (Shilluk) anda team was expected to work in the other (Noru), withim twelve months. One or two toleB languages were beingconsidered as possibilities tor joiningwith Shilluk and Noruto make up athird cycle of language projects.

Am JUL teem was working on preparing Arabic transitionmaterials in seeperatiem with members of the INL staff. These materials were to be tested issohoole beginning lathe 1984 school year. An SILmember was expected In 1914 to tap op workon the Inglish transition materials.

Bevialems were being undertaken ofthe Cycle Oneprimersand itwas expected that theme would be completedtogetherwith supplementary materials before the end ot 1963 ready for the 1984school year.

Flame esistedtOr cooperating with the CurriculumDevelopment Center in preparieg textbooks in the vernaculars for arithmetic, science and religion. It was enacted that teams would commence translating the textbooks into Role A langmeges toward the end of 1983.

All these plans and most of theactivitiesof the Institute of Regional Lamps.* smd the looal languages literacyproject have, however, been placed la jeopardy by recent politicalevents. In May 1983, theSouthern Regional goverment vas dissolved and replaced bythree new regions--Equatoria, Bahr el Cassel amd Ilpper Mile, each with itsown government. This decision threw the government administration of the South intochaos, and it isstill not clW whether the Institute tor Regional Languages will stirvive, and ifso in what farm.

203 PAGE 112

Rebel activity has also increased to thepoint where all the activities connected with the project in the Upper Nile and Bahrel Ghazal regions have effectively ceased. How soon they can be resumedis uncertain. In general then, until there is greater politicalstability the Institute of Regional in the South, Languages and the local languages literacy project are not able to !Unction noraally. It is to be hopedthat this affairs will soon change and that state of theproject will be able to continue. this proves possible, As the lessons learned inthe first years of should strengthen the project its futuredevelopment and its impactin the Southern Sudan. schools of

John Hollman and John Bendor-Samuel Summer Institute of Linguistics

204 PAGE 113

7.0REFERENCES

Atkins, J. (no date) Educational systems in Maridi, South Sudan. manuscript. Unpublished

Barrera-Vasquez, A. (1953) The Tarascan projectin Mexico. In The use of vernacular languages in education.Paris: UNESCO. Beshir, M. O. (1974) Educational policyand the employment problem Southern Sudan. in Khartoum: University ofKhartoum.

Biemiller, A. (1970) The development of the use of graphic andcontextual information as children learn to read. Reading Research Quarterly, 75-76. 6,

Bloomfield, L. (1942) Linguistics andreading.Elementary English Review, 12, 125-130, 183-186.

Bloomfield, L., & Barnhart, C. L. (1963) Let's read. Bronxville, NY: Clarence L. Barnhart.

Bokaaba, E. G. (1984) French colonial language policy in Africaand its legacies. In K. Williamson (Ed.),Language in African education. Amsterdam: John Benjamins.

Bokaaba, E. G. (1983) Language and literacyin West Africa. In R. B. Kaplan (Ed.), Annual review ol applied linguistics: Vol. 4.Literacy issues. Rowley, MA: Newbury House.

Bowen, D. J. (1968) The UCLA- Phillippines language program 1957-1966. Department of English, Universityof California, Los Angeles. Brosnahan, L. F. (1963) Some historicalcases of language imposition. In J. Spencer (Ed.), Language in Africa. London: Cambridge UniversityPress. Brown, R., & McNeill, D. (1960) The 'tip of thetongue' phenomenon. Journal of Verbal Learning and VerbalBehavior, 5, 325-337. Burns, D. (1968) Bilingual educationin the Andes of Peru. In J. A. Fishaan, C. A. Ferguson, & J. Das Gupta (Eds.), Language.problemsin developing nations.New York: Wiley.

Carterette, E. C., & Jonei, M. H. (1963) Redundancy inchildren's texts. WEISEL-1112.1. 1309-1311.

Chomsky, N. (1970) Phonology and reading. In H. Levin and J. P.Williams (Eds.), Basic studieson reading. Cambridge, MA: MIT Press.

Clay, M. M. (1968) A syntactic analysis of reading errors.Journal of Verbal Learning and Verbal Behavior434-438.

205 PAGE 114

Cohen, A., Frier, V., & Flores, M. (1973) The Culver City Spanishimmersion program: End of Year No. 1 and Year No. 2.Department of English, University of California, Los Angeles.

Cohen, A. S. (1974) Oral reading errors of firstgrade children taught bya code emphasis approach. Reading Research Quarterly, 10, 616-650.

Cowan, J R., & Cziko, G. A. (1908, February) Evaluation ofthe Local Languages Literacy Project in the Southern Sudan: Materials critiqueand impact evaluation plans. Unpublished manuscript, Center forApplied Linguistics, Washington, D.C.

Cowan, J R., & Harmed, Z. (1976) Reading performance ofbilingual children according type of school and home language.Language Learning, 26, 353-376.

rowan, J R. (1980, December) Reporton a reading workshop held for SIL/IRL rsonnel in Maridi: October 1980. Unpublished manuscript, Center for Applied Linguistics, Washington, D.C.

Cziko, G. A., & Cowan, J R. (1982, March) Evaluation ofthe local languages literacy project in the Southern Sudan:Second impact evaluation. Unpublished manuscript, Center for AppliedLinguistics, Washington, D.C. Demos, A. (1984, November) Languageand literacy in Africa: Lessonsfrom Ethiopia. Paper presented at the Round Tableon Language, Literacy, and Communication in Africa, University ofIllinois, Urbana.

Dressler, W. U. (1982) Accelerationretardation, and reversal in language decay? In R. L. Cooper (Ed.), Lailguagea read: Studies in diffusion and social change.Bloomington: Indiana University Press.

Elder, R. D. A. (1966) A comparison of the oralreading of groups of Scottish and American children. Unpublished doctoral dissertation,Department of Education, University of Michigan.

Evans-Pritchard, E. E. (1971) The Azande: History andpolitical institutions. Oxford: The Clarendon Press.

Fishman, J. A. (1982) Attracting a followingto high culture functions fora language of everyday life In R. L.Cooper (Ed.), Language spread: Studies in diffusion and social change. Bloomington: Indiana University Press.

Fishman, J. A. (1971) National languages andlanguages of wider communication in the developing nations. In W. H. Whitely (Ed.), Languageuse and social change: Problems of multilingualism with specialreference to Eastern Africa. London: Oxford University Press.

Fries, C. C. (1963) Linguistics and reading. New York: Holt, Rinehart & Winston.

Fries, C. C., Fries, A. C., Wilson,R. G., & Rudolf, M. F. (1965) To teach reading: The transfer stage. A manualand guide for a basic reading series

206 PAGE 115

developed on linguisticprinciples (A revised experimental edition). Ann Arbor: University ofMichigan.

Gabjanda, J. D., & Bell, H. (1979) Language and educationin the Sudan: Present trends. African Languages, 5 (2),28-32.

Garvey-Williams, F. H., & Mills,L.R. (1976) Education in the England: ILO & UNESCO. Southern Sudan.

Gelb, I. J. (1952) A study ofwriting. University of Chicago Press. Gibson, E. J. (1965) Learning toread. Science, 148, 1066-1072. 4 Gibson, E. J., & Levin, R. (1975) The psychologycf reading. MIT Press. Cambridge, MA.:

Glenn, C. G. (1979) The role of episodic and of story length inchildren's recall of simple stories. Journal of Verbal Learning 1/1 229-247. and Verbal Behavior,

Goodman, K. S. (1967) Reading: A psycholinguIstic guessinggane. Journal of the Reading Specialist, 6.126-135. Gough, P. B. (1972) One second ofreading. In J. F. Kavanagh & Mattingly (Eds.), Language I. G. by ear and eye. Cambridge, MA: MIT Press. Gudsehinsky, S. C. (1973) A manual ofliteracy for preliterate Ukarumpa, Papau New Guinea: peoples. Summer Institute ofLinguistics. Hoppe, R H. (1984, May) Socialconttxt of the Local project in the Southern Languages Literacy Sudan. Unpublished manuscript, Linguistics, Dallas. Summer Institute of

Kahane, H., & Kahane, R. (1979) Decline andsurvival of western prestige languages. Language, 55, 183-198.

Kieras, D. E. (1978) Good and badstructure in simple apparent theme, reading time, paragraphs: Effectson and recall.Journal of Verbal Learning Verbal Behavior, 17, 13-28.) and

Lambert, V. E., & Tudker,G. R. (1972) Bilingualeducation: The St. Lambert experiment. Rowley, MA: Newbury House.

Leap, V. L. (1981) American Indianlanguages. In C. A. Ferguson & Heath (Eds.), Language S. B. in the U.S.A. Cambridge: Cambridge Press. University

Lee, E. V. (1982) Literacy primers: The Gudschinskymethod. Institute of Linguistics. Dallas: Summer

Leu, D. J., Jr. (1982) Oral readingerror analysis: A critical research and application.) review of Reading Research Quarterly,17 (3), 420-437.

207 PAGE 116

Lewis, E. G. (1982) Mbvements and agents oflanguage spread: Wales and Soviet Union compared. the In R. L. Cooper (Ed.), Languagespread: Studies in diffusion and social change. Bloomington: Indiana UniversityPress.

Macnamara, J. (1970) Successes and failurein the movement for the restoration of Irish. In J. Rubin & B. H. Jernudd (Eds.),Can language be planned? Sociolinguistic theory and practice for developingnations. Honolulu: The University of HawaiiPress.

Nahmud, U. A. (1983) Arabic in the Southern Sudan: History and spreadof a pidgin-creole.Khartoum: FAL Advertising & Printing.

Malherbe, E. G. (1946) The bilingual school:A study of bilingualism in Africa. London: Longaans, Green.

Handler, J. M., & Johnson, N. S. (1977) Remembrance ofthings parsed: Story strucure and recall. Cognitive Psychology, 9, 119-151.

Handler, J. N., Scribner, S., Cole, M., & DeForest, M. (1979) Cross-cultural invarianca in story recall. Child Development, 50, 886-889.

McConkle, G. V., Zola, D., Blanchard, H. E., & Wolverton, G. (1982) Perceiving words during reading:Lack of facilitation from exposure. prior peripheral Perception and Psychophysics,3, 217-281. Miller, G. A. (1956) The magical numberseven, plus or minus two: Some limits on our capacity for processing information. Psychological Review, 63, 81-97.

Minium, E. V. (1978) Statistical reasoningin psychology and education. New York: Wiley.

Nondiano, N. (1966) Reading com rehensionin the national language: A comparative study of bilingual and all - Spanish approaches toreading instruction in selected Indian schools In the highlands ofChiapas, Mexico. Unpublished doctoral dissertation,New York University.

Neville, P. H., & Pugh, A. K. (1976) Context in readingand listening: Variations In approach to clozetasks. Reading Research Quarterly, 12, 13-31.

Nezworsky, T., Stein, N. L., A Trabasso, T. (1982) Story structureversus content in Children's recall. Journal of Verbal Learning andVerbal Behavior, 21 190-200.

Preck, R. P. (1973) Realism of self-appraisaland school achievement in eight. countries. In Determinants of behavioraldevelopment. New York: Academic Press.

Ramos, M., Aguilar, J., & Sibayan,B. (1967) The determination and implementation of language policy. Quezon City, Phillipines:Phoenix Press.

208 PAGE 117

Rubin, J. (1968) Language and educationin Paraguay. In J. A. Fishman, C. A. Ferguson & J. Das Gupta (Eds.), Languageproblems of developing nations. New York: Wiley.

Rumelhart, D. E. (1975) Notes on a schema forstories. In D. G. Bobrow & A. M. Collins (Eds.), Representation and understanding: Studies incognitive science.

Rumelhart, D. E. (1977) Understanding andsummarizing brief stories. In D. LaBerge & S. J. Saauels (Eds.), Basic processes in reading: Perception and comprehension.,

Ryan, E. B., & Semmel, E. B. (1969) Reading 83a constructive language process. Reading ResearCh Quarterly, 5, 59-83.

Scharrer, G. (1983, May) Southern educationlanguage barrier. 27, 32. Sudanow, pp.

Shannon, C. L. Heaver, N. (1949) The mathematicaltheory of communication. Urbana: The University of IllinoisPress.

Smith, F. (1971) Understanding reading:A psycholinguistic analysis of reading and learning to read. New York: Holt, Rinehart &Winston. Spolsky, B. (1972) Attitudinal aspects ofsecond language learning. In H. Allen & R. Campbell (Eds.), Teaching English as a second language. New York: McGraw-Hill.

Stein, N. L., & (1979) How childrenunderstand stories: A developmental analysis. In L. Katz (Ed.), Current topicsin early childhood education (Vol. 2). Hillsdale, NJ: Ablex.

Stein, N. L. Glenn, C. G. (1979) An analysis of storycomprehension in elementary school children. In R. Freedle (Ed.), New directionsin discourse processing (Vol. 2).Hillsdale, NJ: Ablex.

Stevens, K., & Halle, M. (1967) Remarks on analysisby synthesis and distinctive features. In W. Valthen-Dunn (Ed.),Models for the perception of speech and visual form.Cambridge, MA: MIT Press.

Thorndyke, P. V. (1977) Cognitive structures incompreheasion and memory of narrative discourse. Cognitive PsychologV, 68, 643-852.

UNESCO (1953) African languages and.English in education (EducationStudies and Documents No. 11). Paris: Author.

Valencia, A. (1970) Bilingual/biculturaleducation: An effective learning scheme for first grade Spanish- speaking, English- speaking, and American Indian - speaking children in New Mexico: A report ofstatistical findings and recommendations for thegrants bilingual project. Albuquerque, NM: Southwest CooperativeEducational Laboratory.)

Valencia, A. (1970) The relative effects ofearly Spanish language instruction on Spanish and English linguistic development: An evaluationof

209 PAGE 118

the Pecos Language Arts Program for the Western States SmallSchool Project., Albuquerque, NH: Southwest Cooperative EducationalLaboratory. Kintscb, V., & Van Dijk, T. A. (1975) Comment on se rappelle eton resume des histoires.Languages, 40, 98-116.

Venezky, R. L. (1970) Regularity in readingand spelling. In H. Levin & J. P. Williams (Eds.), Basicstudies in reading. New York: Basic Books. Veber, R. M. (1970) A linguistic analysisof first-grade readingerrors. Reading Research Quarterly,5, 427-451.

Weber, R. H. (1970) First-graders'use of grammatical context in reading. H. Levin & J. P. Williams In (Eds.), Basic studies in reading:.New York: Basic Books.

210 PAGE 119

8.0APPENDIXES Appendix 3A 120

(cont'd.)

212 Appendix 3A(cont'd) 121

Bungaya kiya

Bungaya gina kiya. Tindo gina lamba.

213 Appendix 38 122

1 se

se e e e se tene

se se tene te satana ne sitini

se nete sene -tene Nakoco ka kitito. ka kitito kitito

2 ka kitito Nakoco ka kitito. Bungaya ka si se. A Tindo ka nono sino. Nakoco ka kitito. Nakoco ku kitito.

215 i A , I 2 2411 I i a I * 4 i I LI a I r. t*

so 0 a.. . . , II I i ) t 1 TASHAk .41 IIRGIN KAt.A .4,I'.s "s r"Ps I. 1 41

I You Myud'aF1'adimatu,:tlaiesda',

IWowsun to toshar prgIn route din , 10 ,mopjul 0 toilerMIA kasa010 Aliyudo Faciimatuyanzul tuktoto tit ,o 4akayansu. I. Ga su nan cIknmota zaune. ..' , .. I Suna I Gp.Alliu diFatitioatusunaduba:, "11 zaune a kan kayansu. . , . ,, ' IflQZQSUtQf ii .. ,.. s1 I ' . s Za sü tafi 1 , hutu . I 1' tare da kawunsu. s ! 4, ,.41 e 1 I ,I Ina kawunsul

I 1 . ke SI I' 4 "0.' :tilt Yana I ,: I: zaune clIdn motcia gaba. 0 . 5 lit. I l'a '9 .1. 'aS . t° t .21i S. BEST COMmaw

218 126 APPENDIX 3F

New Format for Latter Drills Proposed by Joycelyn Clavinger.

J cibat

3. a Ca CO

3 Ioi Ca CO cik cak cal cik cak kak VI frai kol ligmew

5 ... . .

. 441111-ailVAti... 46#*, cibat c± cieq cazabak

219 APPENDIX 3G 127

Directions for Writing Conteict CluesDrills

Lessons 1-25

1. AlaUks should be near the end of thesentence.

2. The word you leave out of thesentence can be any type of word: noun, verb, adjective, functor, etc.

3. The 3 word which youlwrite below thesentence should be different Linde of words: nouns, adj. functor., verb, eta.

Lessons 26-50

1. The blaMk can naw be in the middle of thesentence as Well as at the end of.the sentence.

2. The word you laza out of the.sentencecan be any t7pe of word..

3.The 3 wordJiich you write below thesentence should be the sand kind of words with only one of them beingthe correct word.

Lesson 51-

1.. The blank may be any place in thesentence: beginning, middle, end.

2. The word you leave out can be.any tYpe of word.

3. The 3 words which you write below thesentence should look ver- much alike.

Directions for Teaching Context Clues

Nave the pupils open their books to correct page.

2. Point out that a word is missing in each of thc.sentences.

3.. Asi the pupils to read silently the1st sentence and the words under the sentence, and to choose the best wordthat completes the sentence.

4. Ask the pupils which word is thebest word to complete the sentence.

5. Ask the pupils to read silently the 2ndsentence and the words underneath, and choose the best word that completes thesentence.

6. Ask the pupils which word is the bestword to complete the sentence.

7. Continue read each sentence and the wordsunder the sentence until all the sentences havelmen read und the best word chosen for euchsentence. APPENDIX 3H 128 Proposed modificaticn of Gudschinsky method functor drills, especially for functors with illustratable usages. Data from Xresh.

by Rick 3rown

Drill 1

picture of dkcl 'Koko is walking.' boy walking Xrersir6nt is milking Ufa walk

Comments on drill 1: The-pupils are told that the sentence describes what is happening In the picture. They are instructed to look at the picture and then try to read the first line.The picture should help chem to attack or guess at a and ni. They should already knew ad and lb. The teacher can then teed the sentence to them correctly and point out to them the. new functors. Lines two and three could be optional, as could the underliningor circling of the new functor (s), but one or the other should probably he used, if not both. The letters of the new functots may.or 2:nry not be known to the pupils.The primary modification is that an illustration !,4; supplied to pro- videclues to tho pupils to help them to read the sentence before it is read to them by the teacher. By beginning with the meaning, the drill encourages .the pupils to decode rather thatn recode the-words of the sentence,

Drill 2 picture of les761 'walk/hc-walked' boy walking, Ar1646 'Be-is walkieg.' continued r\qi. 1646 'Re-is still wal7ting.'

picture of restie wazeye 'Re-shelled ground-nuts.' 'boy X Asa 'Re-is ahalling ground-nuts,.' shelling XVisa ur 'Re-is shelling them (VP-final mark groimd-nuts 7 tal rts' nr I lie-is still shellingthem '.

Comments on drill!2:This .drill could be optional. When used, it could be :more of a transformation drill, changing toot forms into other foens, or at leas citagiug the known into the unkown.This could include the addition of morphemes rioftbonrse.The purposes of this drill would continue to lucluda chunkintend focuH on exhibiting the function of the new functors, but in addition to this, the VaTIOUS fuuctions of a single functor could be demonstrated, contextually trigge: morphophcnemic Changes (transformations) could be exemplified, and pictures coul assist-the pupils to decode, Wthe Xresh imperfect verb phrases, ni is raquira to mark the end of che verb phrase ii there is no marked direct object or advert in that positicn (the inanimate Airect object is s zero morpheme.)Thus ni occurs quite expectedly after at intransitive verb like love 'bathe' in an im- perfect verb phrase. It recurrs when vazeye is replaced by a pronoun)g 'it'. The contextually sensitive morphoronic change of I to rbefore t is practiced twice. The following third build-up could have been provided: CJ

I

Basple lievision Lesson.

The sample on the oppositepage is based cm'

Lesson6 items taught in lessoni.1-5of the current Zande primer nonuser*.

Items.to bereviewed:.

Ione; introduced in lessons1-5: a mgbe gb,r,a ondI. labia ngbe trot lunotors taught $.n le8eon9-5:

ki "then' (vetbaspeot norver)

a . 'past tense barker'

,Verds used in st?riesof lessons 1r51 hie a b4i, baga. Nouns: aimpu 'beans' big', obey% b4ga 'basket' ga 'hoe' 'mat'

Verbst ga 'return' k bj 'see' Inie ki hiAmmo: 'eat' 4

tl ri App lther known iitornatioh RI (words taugbt inprowpriner 10

Itnetorsi ku.p.yo 'in' Ri,.. 'she..,'

haulm Dobira 'Name erinie's labia4go k%kporo Yo fathel Siryr oldriirl lebira a ba Iga ku kporo 704.. kyoro ,'home/homestealis

Yerbe ngbe bought

(cont'd.) 222 223 Appendix 31 (cont.d) 130 Procedures for Writing Revision Lessons by Julie Van Dyken

Step One: List items taught in lessons since the last review lessons: letters lunctors

. Content words

Step Two: Identify .substitutable words. Relist the words above (Product of Step One) according to semantic and/or syntactic categories SUbject sauna Object nouns Verbs Other

Step Three: Identify words which are minimal pairs or homonyms. e.g. bi ga bi ba ba . ri

Identify other "look alike" words"

baga bagi

MR: In early lessons there may not be any:minimal pairs to contrast.

Not all languages have monosyllabicwords for contrastive purpcees. A, single syllable may never have meaning alone. If such pairs do not exist, then it is questionable why-a reader should be asked .to learn the minimal contrast anyway.

Step Tour: Select which words for review (listedunder Steps One through Three) need focus.These words should appear in the substitutionslot of the sentences accompanying the pictures.

Suggested guideline for selection priorities could be 1. Tuictors a. Graphically siiilar words 3. Lets common contentives substitutes 'Step live: Wite sentence pairs in which the second sentence one or wore words used inthe first.

2be sentence pairs should substitute 1. graphically similar words 2. functors.(if the difference is picturable) 3.words which are both graphically andsemantically

dissimilar. . (cont'd.)

224 Appendix 31 (cont'd.) 131

Step Six: If desirable and possible, order the sentence pairs ina logical order. e.g. Dabira buys a mat in one picture. In a later sentence pair Dabira throve the mat into the house.

Instructions to the teacher: Rave the pupils identify the sentence that says vhat the picture says: "Look at the picture. Read the two sentences.Which one says what the picture says?" NB: Ihe'lesson plans must require each pupil to read both sentences, so that words in the lwrong" sentence are a/so revieved.

Problems vith the lesson type: a. Ninimal pairs may fill different syntactic positions. b. Functor differences may not be picturable.

. '225 APPENDIX 4A 132 Zande Literacy Test 1. Nonsense Syllables ga, ma, ra, bi, ru,gu, ra, ma, mi, ba, gä, IA 2. Known Words

kporo 'house nvugg 'monkey' nzengu 'market/guinea fowl' begi 'sold' gita 'hoe' bundu 'gun' abapu 'beans' suma 'fish net' ira 'mat' bibiri 'black' 3. Unknown Words

meme 'goat' iga"hide' gama 'give away' tita 'grandfather/mother' ora 'run' nzeme 'fat' onga 'breath' Dari 'frog' ugu 'dry' bagara 'cow' 4. Reading Comprehension

Passage

Nako a ndu ku Bako yo, I ki ndu na Bako (name) past go to (name) there they then go with (name) ku nzengu yo ku ngbe abapu na boko. I a ndu to market thereto buy beans and pumpkin They past go ki ngbe abapu na boko, ki ga ku kporo yo. then buy beans and pumpk4h then return to house there. i al ga ni, i a bi mama. When they progressive return (when ) they past see leopard It aS ri ango. He i a bi mama ni, i progressive eat dog When they past see leopard when they a oro be ru. Ho i al ora ni, abapu a past run from it When they progressive run when beans past xa be Nako na boko a ti be Bako. I ki pour from (name) and pumpkin past fall from (name) They then oro ku kporo yo. run to house there 226 (cont'd.) Appendix 4A (cont'd.) 1.33 gliak0 went to Bako's place, then they both went to the market to buy beansand pumpkin. They went and bought some beans and pumpkin and then returned home. On their way (as they were returning) they saw a leopard eating a dog. As soon as they saw the leopard, they ran from it. As they were runningaway, Nako dropped the beans and Bako let go of the pumpkin. They ran (all the way) home.'

Comprehension Questions

1. A da na a ndu nzengu yo? pl. who rel. past go market there 'Who went to the market?'

2. Gini i a ndu ka mange ha nzengu yo? What theypast go in order to do it market there 'What did they go to do in the market?'

3. I a ndu ku wdri fuo ngbe yo ahe nzengu yo? They past go in order to where after buy their things market there 'Where did they go after making theirpurchases at the market?'

Gini hl i a bi h6 i afi ga ni? What thingtheypast see whenthey prog. return when 'What did they see as theywere returning?'

mama ad ri he? What thingleopardprog. eat it 'What was the leopard eating?'

6. Tipagine i a oro tini be mama? Why theypast run why fromleopard 'Why did they run from the leopard?'

7. Gini pai i a magi he na agu ah6 What thing they past do it with thosethings

a ngbe he nzengu yo? they past buy then m,.rket there 'What did they do withthe things they bought at the market?'

227 , APPENDIX la 134

Nuer Literacy Test: Primary 1 and Beginning Primary2

1. Nonsense Syllables go, ko, lo, ga, ka, la

2. Known Words

lak 'to wash' moc 'to give a gift' lath 'to put' thal 'to cook, boil' kal 'compound, fence, child' bal 'to roast' gat 'child' bith 'fishing spear' gook 'baboon' joc 'to chase' 3. Unknown Words

thlk 'steer' wic 'village' lok 'floating free, blowing in hak 'to split open' the wind' mitt) 'food, to eat intrans.' yak 'hyena' ben 'to come' geth 'to clean a barn' cak 'milk' .raan 'person'

Reading Comprehension

Passage

Man DEEg cc wl rey kal. Gat Man DEE0 Mother of DEE!) she/p4st go intocompound. Child (of) Mother of DEE0 a thin. Deg kEnE Jok cikE wS thin bS. man DEEg is present. (name) and (name) they/past go therealso. Mbther of DEEg cc Deg kam bith. CikE wl nay yieer ke she/past(name) give fishing spear. They/past go .fishing(in) river with

.bith. J3k cc rec nlk. Cike je bul. Ci DEo fishing spear. (Name) he/past fish kill. They/past it roast. He/past (name) rec nk Cike rec DEE0 nag. Cike via kakS. fish kill also. They/past fish (of) DEEg take. They/pest go (to) field, Gook a rey kakS. Jok ce gook joc kakS ke baboon is in field. (Name) he/past baboon chase (from) field with th DEEg. Cike rig wic ke rec. spear (of)DEEg. They/pastrun viilage with fish.

(cont'd.)

228 4 Appendix 4B (cont'd.) 135 'Deo's mother went into the compound. Her child was there. Deg and Jok also went into the compound. Dee's mother gave Dega fishing spear. They (Der) and Jok) went fishing with thespear. Jok killed a fish (with it) andthey roasted it. Then Deg killed a fish too. They took Dee's fish andwent into the field. There was a monkey (a baboon) in the field. Jok chased the monkeyfrom the field with Dee's spear. Then they ran to the villagewith the fish.'

Comprehension Questions

1. Kee nhibm Deg kene Jok cike wN ni? at first (name) and (name) they/past go where 'Where did Dee and Jokgo first?' (Answer: into the compound)

2. e oa te thin rey kal? It was who waspresent in compound 'Who was in the compound?'

3. Cike eu lel ke rec Joak? they/past what do with fish (of) Jok 'What did they-do with Jok'sfish?' v. Cike. eu lel ke tec DEW They/past what 'db with fish of Deg 'What did they do withDee's fish?'

5. Cike eu lel my kakti? They/past what do in field 'What did they do in thefield?

6. c ou lat ken e 0? It is why do they it why 'Why did they do it?'

7. c ou caa rec DEEr) nag wic 6? It is whypassive fishof Deg bringvillage why 'Why did they bring Dee'sfish to the village?' 'Why was Dee's fishbrought to the village?' WNW so

tow teem* low $1010011101110 OirisePt

Vs% 'OW III OW' 00 11, el000 solo' 1110111101fesor /IS oosroloo IOW loon %A* 14110411 NSW Olt 'oitoofois*

IN four lb owe 41 pers ist MOM sow 10 41011110110 MPS IOW hist *Or pee

sol, hi Os gooll hi) co Jot OWN SW torril 4110*** si faMor of lift. (NN)bahest (1181) a. ci iftelt at be I. M cc OPOIASNO 110 tog. Olinal Waist 00leo roeorm. (a) boost oge Ili es OWL Os es we MO IL CM to Otot a* 110 0.01. IIIEW Won ftoow ss. IhRepest tha 11. ell WiS MO. as to ad WI. elks ell ells' WI Or* Illeshistfist On WNW Nast.Iheyftsst Oaa el as all* he mit MI. OM roc altillr ri islinst MO tat ON isososiod. low/Postfish eh sr leo Mb. Illio 50/1 ONO fe ear sI 110.Ihsfost now se flo ft taco. On lotei elle el sot lei. agi 01111.11 OM IMO IMAM 0610 1101e (IQ Wu.It ls aio oft .0114.1 Isam eito rey 094asure. anemmer,Poet or (m) ihrmastp I. mem UMW On NW Nam 230 Appendix 4C (cont'd.) 137 'Deo and Jak are Min Deco's children. Dv) gave Jok a fish hookand he baited it with a frog. Jak taught a fish with the fish hookand Oct) also caughtone. Ihey killed them with Dco's fishing spear. Then they roasted Jak's fishand after- wards returned to the village. Jak put the fish hook back in thecompound. They gave Oco's fish to Man Ikea and she cooked it.

Oil and Jak took a sheep to the barn. The sheep belongs to Man Dcqand her chidren. Oco and Jok th,m1 returned to theirmother's compound (having safely secured the sheep in the barn).'

Comprehension Maim

I. ou mi ci Jak c lath kV cibat? It was what rel. he/past (name) it put on fish hook 'What did Jak bait his fish hookwith?'

1. Caa rec nVk kc ou? Passivefish kill with what 'What were the fish killedwith?'

3. Cikc ou lel kc rec Joak? They/pl. what do with fish (of)Jak 'What did they do with Jak's fish?'

4. Cikc ou lel kc rec DEfq? They/pl.what do withfish of Dco 'What did they do with Dco's fish?'

S. c oU mi kam kin 'roc Man Dece It is what rel. give they/pl. fish mother(of)Dco 'Why did they give the fish to Mankeg?'

6. c ou id cu kin c nao luaak? It is what rel. narr. they/pl. it take (to) barn 'What did they take to the barn?'

7. Man a ou dicn? Mother of Dco she is what their 'Man OM is their what (what relationship)?'

231 APPENDIX 4D 138 Nuer Literacy Test: Advanced Primary 2

1. Known Words

rieyde 'canoe/his' waregakni 'papers' waath 'kind of reed/ dhool 'boys' river access' guandien 'father(s)/their' duoor 'thing' eloo 'very' locdien 'heart/their' gnile 'he is looking at/ cl 'Nile perch' . checking uo on' 2. Unknown Words

Kuanylualthoan 'place name' runwao 'morning' Torpuot 'place name' tuytuy 'water hyacinth' Waakow 'place name' jiök 'to say' Wunkir 'place name' kor 'stream' cieo 'home/to live' thief) 'to fill' 3. Readino Comprehension

Passaoe

Mee dan tee dhool mi coali Ruot. Dhool emo Sometime ago there was boy rel. called(name) Boy that te dhoorien kN Torpuot. Mee caa duelgorkN Kuanylualthoan is present home/their at village name.When passive school place name Iep kN Wunkir, CUE ben rtek ke 06 be wN dueelgorkE. open at place name he/nan. come town in order to he fut. go school. Jen cua je nhok e kali* dufflgorka, ki ke kuty He pass. nan. him accept by head (of) school but for sake of dafl boa thile guaath mi ce jek. Cue house of sleeping there was no place rel. (which) he/past find. He/narr. loc dhoorien kl Torpuot. return home/their at place name.

Cue je lar guan en y66 caa nhok duelgorkS kN He/narr. it tell to his father that passive accept school but thiele guaath tooca ke kutyde. Cu guan there was no place of sleeping for sake/his. Narrativefather/his him jiök, "Jin bi te cieo waneme. Thile mi say, You you/fut. be present home here. There is nothing which

(cont.d) Appendix 4D (cont'd.) 139

bi lat. Bi duodu a luoth ke runwaoni you/fut. do/work. You/fut. belonging/your hab. .getting off early inmornings diaal, wii dueelgorka ka kuanylualthpan." all you fut. school at place name pres.Bo

Ke kuiy eft la Rupt a jien ka Torpuot For sake that habit. asp.(name) habitual aspect leave from place name ke runwaq ka bee Waakow ben ook. at morning and he/fut. name ofa stream come cut/cross.

Mee ci nhiaal ku twit dim, ci kor Waakow thiao. When asp. rain begins fall, asp. past stream name ofstream fill. Mee /ken Ruot riey jek, ke y63 e dhool mi pel When he neg. past (name) canoe find, because he boy rel. who clever elan, be tuytuy dol, be bokni kens bieeyni ke very he/fut. water hyacinths gather, he/will books and clothes/pl. them lath wic toytoy. BE ben kuic kete, poke put on head of water hyacinth. He/fut. come across swimming pres. pushing tuytuy. Jen Ice duelgorkl pal. Ka thaao water hyacinth. He neg. pres. school miss/be absent. But some guaathni be cop ke jpak. times he/fut. arrive late.

Mee ci kuaar dueflgorkä je thiec ke la bee Whenasp. pastchief (of) school him ask why habitual come kE jpak, CUE je lar i e nan dhpara ka Torpuot. Cu late he/narr. it say that is far home/my °at place name. He/narr. kuNar wic de car ka CUE Rupt moc guaath toad chief head/his thinkand he/narrative past (name) give place of sleeping dueelgorka ke yera thile ram mi mar ke reek. (at) schoolbecause there is no person rel..whorelatedhim(at) the town. /Ken Ruot le luoc dhporien, cue lari guan He/neg. past (name) again return home/their,he/narr. past told father/his lan CC guaath boa jek duielgorka. KW cu guan message he/past place of.sleeping find(at) school. And he/narr. father/hi (coned.) Appendix 4D (cont'd.) 140 lscde teeth els!) ke yä ci gatde guaath jek. heart/his fall/happy very because he/past child/his place find. 'Once there was a boy called Runt who lived in the villageof Tsrpust.When the (boarding) school Kuanylealthusan openedup at Wunkinhe went there to enroll. He was accepted at the school by the headmaster butunfortunately therewere no more beds in the dormitory left. So Rust went back to Tsrpust.

He told his father that hehad been accepted at the school,but thatthere were no boarding spaces available. His father then saidto him, "You will stay here at home, but you won'tdo any work. Every day you will getup early and go to the school at KyanyluElthusan."

So every day Rust left Tsrpust very early in themorning and crossed Waaksw creek. However, when the rains began to fallsthe Waakswbegan to rise (toa height whichwas to deep to ford on foot). When Rust couldn't finda canoe to cross the stream with, he, being a clever fellow, would puthis books on a thick pad of water hyacinth andswim across pushing the hyacinth ahead of him. He never missed a day at schoolobuthe sometimes arrived late.

When the headmaster asked him why he oftencame late to school,Rust said, bOecause my home at Tspust is very far from here."Then the headmaster decided to.give Rust a place to sleep at the school because therewas no one related to him in tydn who mightput him up.

Rust didn't go back to his house, rather he sent hisfather a message saying that he had found a placeto stay at the school. His father wasvery happy to hear that his son had founda place to stay.' Comrehension Questions

1. Runt te dhsarien winith? 'Where does Rust live?'

2. t gu mi la rick Rust? 'What is troubling Rust?'

3. t gu mi caar guan Rut c ke kuiyw gatde dueelgsrkl? 'How does Rust's father feel about his son's (child's)going to school?'

4. t guaath ith rey runny mi laRust a IA kä Torpust? 'What season of the year is it when Rust IOS (is)coming from Tsrpust?' 5. La Rust yier a guokidi?

'Had did Rustcross (habitually) the river?'

6. t gu ml moc kc Rustguaath tope/ e Why wis-Rust given a place to sleepby the headmaster?' HcouleristeAhatjhe:boywesVeryserous_about (Answer: because he gp.:,!), school.)

: 141 APPENDIX 4E

Frequency Count of Wordsand Grammatical Morphemes in Lessons 1-15 in Zande Book2 A. Nouns D.Verbs (continued) Dabira Male name 33 ga return/go Inie Female name 39 (awayfrom) 30 Naguru Female name 21 bi see 37 nzengu market 33 baga1 sell 12 baga basket 10 ngbe buy 16 atio fish 7 ra sleep 2 kporo house, home 27 rari dream 2 abapu beans 14 imi kill 7 bSgibägl something for ru stand/be at sale 12 a place 2

Ira woven mat 11 ri ate 8 boko pumpkin 5 aa progressive morpheme 10 gita hoe 10 a past bundu gun 8 morpheme 66 ba father 18 mama leopard 17 E.Adverbs nvugi monkey 12 HV...nlwhen 11 uru during the B.Pronouns day 2 ri she/her 14 ki then 71

ko he/him 11 Yo there also 68 they 16 it 8 F.Conjunctions/Prepositions ku C.Ad'ectives to/toward 45 na and/with 28 bibiri black 10 be from/away from 5 D.Verbs a also 11 ndu go 36

11 have not providedseparate counts for the past formbagi which appears more frequently than the citationform.

235