Basic education amongst national minorities: the case of the Yi in Sichuan Province, China - Keith M. Lewin1

Keith M. Lewin (United Kingdom)

Professor of Education at the University of Sussex. He has worked extensively in Asia over the last twenty years on a range of policy and planning projects. Co-presenter of a round-table at the World Conference on Education for All, Jomtien, 1990. Since 1984 he has directed exchange programmes with Beijing Normal University and Hangzhou University in China where he holds honorary professorships. His books on China include Implementing basic education in China (1994) and Educational innovation in China (1994).

Introduction

This paper offers a descriptive analysis of formal basic education provision amongst the Yi people who live in mountainous areas in the south-west of Sichuan Province in China and some reflections on the problems of educational provision for minority groups. Enrolments in primary school amongst the Yi remain well below universal levels and the drop-out rate, especially amongst girls, is considerable. The levels of achievement of Yi students at the end of primary school are well below those of Chinese pupils living on the plains. The physical condition of schools is poor, and many of these school buildings were originally intended for other purposes. A number of steps have been taken to overcome some of the main problems. These include sanctions against families who withhold children from school, subsidies to reduce the direct costs of schooling, the introduction of girl only classes and special boarding arrangements, and regional incentives for teachers to retain girls in the classroom. This case study concludes with a discussion of some of the questions raised by the analysis.

Background

China has over fifty national minorities mostly located around the periphery of the country and in the more remote and inaccessible parts of the interior. The Yi population live in mountainous areas spread across Yunan, Guizhou and Sichuan Provinces in China. The best estimates suggest that altogether they number about 5 million people of which 1.5 million live in the Liangshan mountains around Zhaojue where this study was based. Zhaojue is amongst the poorest 100 counties in China with an income per capita less than one tenth that of the richest counties. In terms of wealth it is in the middle range of income for autonomous regions. About 30% of rural families are estimated to live in conditions below national minimum standards of nutrition and shelter. The climate is harsh as a result of the altitude (up to 3,900 metres above sea level) and the terrain is rugged, making communication difficult. Towards the end of the 1980s illiteracy rates amongst the Yi were about 46% for males and 78% for females (cited in Kwong, 1989, p. 95).

Yi society before liberation was essentially feudal in character. Traditional leaders controlled the use of the land, and labour was provided by serfs bound to the landowners whose status was inherited. Subsistence farming was the main occupation. Yields were low and methods primitive. Rival clans competed and fought over the small quantities of essential commodities imported from outside the area, such as salt. Since liberation, the State has taken responsibility for the administration of the area and has replaced traditional governing structures with those common in other parts of China, thus creating a local bureaucratic administration which exists alongside a party organization. Agriculture has been modernized at least in part, and livestock and cash crops have become important sources of revenue.

Zhaojue county covers seven districts with a total population of about 200,000 grouped in 47,500 families, of whom over 96% are Yi. Most of the remainder are Han Chinese. Population growth rates officially appear to average around twenty-five per thousand, having fallen from over forty per thousand in the early 1980s. The death rate is reported at approximately eleven per thousand. Males consistently outnumber females in the population by 4% to 5%. The national family planning policy applies in Zhaojue but is not strictly implemented. A farming couple are allowed to have three children. Data from field studies indicate that it is quite common to have more than three children, and that the real rate of population growth is likely to be larger than the official figures indicate (Lewin, et al., 1994b, p. 116).

Like many other national minority areas, the social and cultural environment in Zhaojue is impoverished. Those with any educational qualifications work mainly in the education system. There are very few technically qualified staff, and they are concentrated in the county town and in a few factories. There are only two people with formal engineering qualifications in Zhaojue. Before 1980 almost all of the farming population was illiterate. A literacy campaign was conducted in the early 1980s in the Yi language. Though it is claimed officially that most young and middle aged people can now read the Yi script, there is no independent verification of this and it is probable that some of those who acquired literacy have subsequently lost it. There are virtually no books, journals or broadcasts in Yi. The mass media in Chinese are not easily accessible; there are no television sets and few radios in most areas

Two areas were selected for detailed case studies within Zhaojue. These were Bier and Sikai. Bier district is thirty-two kilometres north of Zhaojue County. The Bier River goes through the district from north to south. The only flat land is along the bank of the river and the rest of the district is mountainous, with an average elevation of 2,500 metres. Amongst the eleven sub-districts three have no radios and another three have only unmetalled roads. Four of the eleven sub-districts have no electricity. The population of 30,000 is almost entirely Yi, with a low population density of only sixty-four people per square kilometre. Bier is almost entirely dependent on agriculture, which is severely constrained by the mountains and climate. It remains a very poor district.

Sikai District was the other case study area chosen. It lies more than twenty kilometres south-west of Zhaojue County town. Several small rivers run across the district and it has a relatively large plain area. On average it is 2,200 metres above sea level, is a little warmer than Bier and has a higher rainfall. Communications are slightly better with more roads, although three of the nine sub-districts have as yet no bus service. Two thirds of the sub-districts have electricity. Like Bier, Sikai is almost entirely populated by Yi people. The total population in 1991 was estimated at 35,400. In Sikai the number of females has only recently begun to approach 50% of the total population.

Sikai is marginally more developed than Bier, but it is still very poor and educationally undeveloped. In 1991 there were only three college graduates and approximately 200 upper secondary graduates in the district. Only 2% of the population had completed lower secondary school. The medium of common communication is the Yi language and few people understand standard Chinese except officials and teachers in both districts.

The school system

Before 1950 there was no formal schooling amongst the Yi in Zhaojue. What education there was consisted of informal learning at the feet of a small number of special members of the community. The master, who was called a bimo, taught his followers, who were mostly drawn from his immediate or extended family. There was no systematic educational provision for other members of the community drawn from lower status groups. Traditional learning was religious, based on customary beliefs, and undertaken only in Yi language. What literacy there was in Yi (estimated at 3% in the 1950s) was concentrated amongst the land owning classes. Though the Yi language has a long history of more than one thousand years, it is an irregular language, and the written form used by the different bimo varied.

The government of Xichang Prefecture established the first primary school in Zhaojue in Sikai District in 1940. It enrolled thirty boys but it soon closed. A new government was established in June 1950 after Liberation with a Department of Education. Three primary schools were set up by the end of the year. Schooling was free and all costs were paid by the government. During the Great Leap Forward (1958), and during the Cultural Revolution in the 1970s, schools were supposed to exist in virtually every village. In reality many were not in special buildings and sustained financial support was not forthcoming.

Today, Zhaojue has two complete secondary schools (with both lower and upper secondary grades) and one lower secondary school which is reserved specifically for Yi students. There is also one teacher in-service training school. All these institutions are located in the county town. The two primary schools in the county town are former provincial key point schools, and the seven central primary schools at district level are all complete schools (grades 1 to 6). Three of these schools offer lower secondary classes. Only ten of the remaining sixty-two xiang (sub-district) level central primary schools are complete. The remaining village schools are widely dispersed and are often shared by more than one community. The typical village school offers only grades 1 and 2.

In Bier, of the thirty-seven primary schools only the district central primary school offers all grades. Most of the eleven xiang central primary schools cover grades 1 to 3. If students who have finished three-year schooling want to continue their education, they must pass entrance examinations and transfer to other districts or county town schools. The twenty-five village schools usually only offer grades 1 and 2. There is one school for each large village, otherwise several villages share schools. The county government has ruled that no school should be established in places where there are fewer than twenty-five students because of the shortage of teachers. As a result there are no schools in the deep mountainous areas.

Sikai has twenty-eight primary schools, including one district central school and eight xiang central schools. Only the district school and three xiang central schools are complete. The nineteen village schools cover grades 1 and 2 as in Bier. There are a few schools in high mountain areas. The local regulations on minimum school size are the same as in Bier. The primary schools in the county town are the largest, with enrolments of more than one thousand pupils. District central primary schools typically have six classes and 200 pupils, though some are larger. Bier central primary school has twelve classes and over 400 pupils. Most village schools have a single classroom (occasionally two) with one teacher and a couple of dozen pupils.

As in most other rural areas in China, the management of the school system is undertaken by principals located in central district primary schools (Lewin, et al., 1994a, p. 202) Separate vice-principals are responsible for instruction in the central school and instruction and administrative work in other schools in the district. Central schools therefore oversee staffing, check attendance records, allocate funds granted monthly by the county Bureau of Culture and Education to the schools in the districts, place and transfer teachers to the different schools, authorize the hiring of substitute teachers, and promote teaching and educational development activities in the district.

Though the physical condition of school buildings has been improved significantly, the quality of what exists remains well below national standards. Village schools are often simply four walls and a roof, and many are converted from other uses, such as cattle sheds. At least 30% of pupils in rural areas lack desks and benches, and we estimate that in village schools more than half the children sit on the floor as there is no furniture. About 300 teachers in the county do not have access to a teacher's room, and 200 do not even have proper beds. There is no physical education equipment, nor other teaching aids of any kind in most of the village schools, and there are no library books.

The immediate reason for the poor conditions is lack of adequate funding. Paradoxically, the average educational expenditure per pupil in the area is not particularly low (partly because Zhaojue has a large proportion of government teachers whose salaries are higher than those of minban and temporary teachers). However, expenditures on personnel account for 96% of the total budget (including the government aid for boarding provision) in Zhaojue, thus only 4% is left for operating purposes. This is not enough to cover office expenditure and is insufficient to support the improvement of conditions in the schools.

ENROLMENTS

Bier had 1,700 pupils enrolled in primary school in late 1991, 14% of whom were girls. Though the total number enrolled was growing, the proportion of girls appears to have been falling. Enrolment decreases rapidly as the grades increase. There were 438 new students for grade 1 in 1984, but only ninety-two were still enrolled when the class reached grade 5 in 1989. Only seventeen were promoted to grade 6. The following year, grade 5 enrolments were even lower.

The highest attrition rate appears to take place between grades 3 and 4—on average grade 4 is 50% smaller than grade 3. The reduction in numbers between other grades is between 25% and 30%. Xiang central schools and village schools usually only enrol grades 1 to 3. Further progression involves a change of schools, which may be both inconvenient and expensive in transportation costs. Places for students in these schools are also limited. About 80% of students enrolled in grade 5 did not progress to grade 6 because the latter grade had only just been introduced.

A similar picture emerges from Sikai, where enrolments decline even more rapidly through the grades. Grade 5 is often less than 20% of grade 1 five years earlier. The highest rate of decline is between grades 1 and 2, where on average grade 2 is about 50% less than grade 1. The proportion of girls is very low and fell to only 11% in 1989/90. In 1990/91, there were no girls at all in grade 6.

ACHIEVEMENT

The available data on levels of educational achievement suggest that performance is very poor. The average scores in both Chinese and mathematics for schools in Zhaojue were lower than the passing scores recorded in the last four years—thus more than 50% were failing according to this definition. Achievement does not appear to have improved significantly over the last five years and there remains a large gap in performance between the town and rural schools. Typically the lowest scores in Chinese language in the town schools are nearly double the average for the rural schools (where Chinese is not spoken outside the school). In mathematics the lowest scores in the town are comparable with the rural averages.

County schools have the highest averages, with the highest number of pupils scoring over 80%. District central primary schools fall in the middle rank, and the xiang central schools have the worst performance. Some xiang central schools in Zhaojue, in more than twenty years of existence, have never had a pupil admitted to a lower secondary school. In the twenty-one xiang central schools, the great majority of pupils score lower than 40% (82% score lower than 40% in Chinese and 95% score lower than 40% in mathematics) in the annual tests. It should be remembered that pupils in these schools are already from selected groups since many have dropped out by the time they reach grades 5 or 6.

The county requirement is that sixth grade pupils taught in the Chinese medium (System 1—see below) should reach a grade 4 level of achievement in the Chinese language in order to graduate from grade 6. Sixth grade pupils in the Yi medium (System 2) should reach grade 4 level in Yi. The acceptable level of performance is set two grades below that of children in schools on the plains to ensure that some are accepted into lower secondary school. Despite these guidelines, teachers at Bier central primary school confirmed that most of their sixth graders in System 1 could not reach grade 3 level in Chinese. Educational achievement in village and xiang schools is even worse. We examined the second grade children in Chinese in System 2 at one of the central primary schools, and found most of these children were unable to read or understand simple Chinese. Class teachers indicated that less than a third would reach the passing level for that grade. Low achievement is inseparable from language problems. The Yi language is very different from Chinese and Yi children live in a Yi speaking environment with no contact with native speakers of Chinese.

Other special measures that reflect these low levels of achievement include concessions for entry to colleges. The admission standard is lowered by twenty points for Han students living in Yi areas, and by forty points for Yi students in order to maintain the proportion of students from underdeveloped areas and minority groups. The Zhaojue government has a similar policy for county secondary school admission.

TEACHERS

In Zhaojue as a whole, there are about 1,000 teachers in primary schools. Most of these teachers are full-time, and about one third are substitute teachers or are minban, who are paid from local funds. The teaching force is predominantly Yi. More than two thirds of all government teachers are Yi, and the minban and substitute teachers are all Yi and are predominantly male. Over 50% of teachers are under thirty years old and are therefore recently qualified.

About 20% of teachers are of senior rank and another 20% are classified in the first rank. These teachers are concentrated in the county town and district central primary schools. There are twenty-six senior teachers in one of the county town primary schools, which is more than half of the total! Unqualified teachers are mostly to be found in the incomplete village schools. In Bier about 70% of teachers reach minimum qualification levels and in Sikai more than 80% which compares well with national averages.

Overall, the pupil-teacher ratio in Zhaojue is claimed to be 13:1, and on average there are 1.9 teachers per class. Neither the teacher-pupil ratio nor the number of teachers per class reaches the level suggested nationally by the State Education Commission. Actual pupil-teacher ratios are likely to be greater since it appears that qualified teachers who are not teaching are often included in the count. Our case study data suggests that in Bier and Sikai, pupil-teacher ratios are about 17:1, which is closer to the national average of approximately 20:1. They have not changed markedly over the last ten years. The average teacher class ratio for Bier was 1.2:1 and for Sikai 1.6:1. These ratios are common in Chinese primary schools—in the case study schools every teacher taught between fifteen and twenty periods out of twenty-five or more. Pupil teacher ratios are lower, and class teacher ratios higher than they would be if multi-grade classes were organized in Zhaojue. It is not unusual for there to be less than ten pupils in each class in upper primary in rural areas but pupils are taught as separately time-tabled groups.

This brief overview draws attention to major educational problems in Zhaojue, many of which are common to other national minority areas (Kwong, 1989). Several points are clear:

— enrolments rapidly attenuate through the primary school system with high drop-out rates;

— completion rates for the primary cycle are very low, and few complete the cycle in the appropriate number of years;

— girls constitute a small proportion of total enrolment (10%-20%) and few succeed in completing primary school;

— class sizes in rural schools may be very small and multi-grade teaching is generally not used, thus costs may be high as a result;

— repetition and over-age enrolment are common;

— levels of achievement are much lower than those in non-minority areas; and the provision and retention of teachers is problematic.

Some interventions

Our research identified seven policy initiatives relating to primary education which are intended to improve access to schools and increase participation rates. These have arisen in the wake of the Compulsory Education Law of 1986 and associated legislation (Lewin, et al., 1994a, p. 33). In Sichuan the Yi and other minority groups benefited from special campaigns to raise awareness of the importance of primary schooling, make greater use of minority languages in schools, and increase the number of minority teachers (Ahmed, et al., 1991, p. 187). In the Yi area a number of specific initiatives have been tried.

First, the xiang level People's Congress has devised local regulations to encourage enrolment. These require that each family send a certain number of its children to school otherwise the family will be fined. In one village in the case study area the rule is that if a family has three school-age children it has to send at least one child to school otherwise it will be fined up to 100 yuan (a third of per capita income). In another village the rule is that 'If a family has two children, it has to send one to school, and if it has three, it has to send two, and so on [...] Farmers who do not send their children to school according to the rules will be fined one mu of land.' Some farmers have been fined but so far no one has forfeited land.

Second, tuition and registration fees have been reduced and exemptions given. A significant number of families are either outside the money economy altogether, or have very low per capita incomes. These families cannot afford the direct costs of schooling. Tuition fees are not officially charged at the primary level, and rural pupils (and those in some district central schools) do not need to pay the registration fee either. Schools should provide textbooks and exercise books for most of their pupils from income derived from school-run businesses. However many schools cannot raise much money since the local economy provides no opportunities for cash generation. Some teachers apparently pay for textbooks for pupils out of their salaries.

Third, special classes for girls have been organized. In order to encourage more girls to attend, some schools have formed all girl classes and have allocated additional resources to these kinds of initiatives. There is some indication that this has a positive effect on enrolment in these schools. An all girls class was established in Bier central primary school in 1987. This school paid for textbooks and exercise books for 60% of the girls in the class; provided a good home-room teacher for general subjects; and selected experienced teachers to teach special subject areas. At the end of 1989-90 school year, all these girls passed the unified county examinations for Chinese and mathematics. Their average scores were around 90%, whilst the county averages were 28% and 49% for the two subjects. The all girls class apparently performed better than other classes in the same school.

Fourth, grants have been made to improve school conditions. State, provincial and prefectural governments have provided significant funding; farmers have donated labour and construction materials. Most of the unsafe buildings have been repaired. Although it is hypothetically possible to generate additional resources from the community to support schools, external government grants are indispensable, especially for the village schools. About one third of the peasants cannot meet their basic needs for food and shelter. None of the schools appear to be generating significant revenue from economic activity. Donations consist of local building materials and labour. Where cash costs are involved, there is usually no obvious way in which loans could be repaid, and grants are usually needed.

Fifth, special efforts have been made to localize the teaching force. In the 1960s, most teachers in Zhaojue came from other areas. A quota was applied to entrance to the four normal (teacher training) schools in the prefecture that guarantees that 65% of admissions each year are reserved for Yi applicants. About 900 graduates from the teacher training institutions are assigned to teaching positions each year in Liangshan Prefecture. Zhaojue receives about fifty new teachers through this route, most of whom were born in the county. As older Han teachers retire and 'go down the mountain' they are being replaced by Yi teachers. The proportion of Yi teachers has increased from only 3 % of the total at the beginning of the 1970s, to over 65% in 1990. Substitute and minban teachers are all Yi, and 90% of the district and xiang central primary school principals are now Yi.

The problem now is to reduce turnover in the teaching force and retain teachers in the area. Zhaojue is not an attractive environment for young teachers coming from other areas since it is both rugged and, for some, culturally unfamiliar. Many of the teachers we interviewed wanted to transfer to areas on the plains or change their job. Bier authority assigns teachers from outside the area to the District central primary school or schools with better conditions along the roads. Those who have local origins are encouraged to work in the semi-mountainous areas and in the high mountains and contribute to the educational development of their home villages. The district authority grants financial aid to local teachers who have children enrolled in the District central primary school or in the primary schools in the county town. Unmarried young teachers are assigned to schools close to each other in order to create opportunities for them to socialize. Government teachers receive a subsidy worth about 25% of their salary over and above what they would earn elsewhere in Sichuan.

The last two initiatives are concerned with the introduction of a boarding system and a bilingual instruction policy. In the 1980s, boarding schools were reestablished and assistance with living costs was provided for Yi pupils. Three types of boarding exist—key-point, general and semi-boarding systems. The key-point boarding system is used for classes of Yi pupils in the county town primary schools; the general boarding system exists in the district central primary schools, and the semi-boarding system is for xiang central primary schools (about two thirds of these offer boarding facilities). Financial assistance varies according to the type of school, and is most generous in the key-point schools. Females receive a higher rate of subsidy than males. Clothing and bedding grants are also given. Boarding is needed since many pupils have long distances to walk to school across mountain ranges.

Pupils are allocated to the different systems according to their examination results. In order to use the grant system effectively, the Zhaojue government has ruled that only Yi pupils at senior grades in primary schools whose families are in rural areas qualify for entrance to the boarding schools. Our evidence suggests that this rule is not always enforced.

The boarding schools differ in orientation. The main purpose of the key-point boarding system is to prepare children for the next level of schooling; the schools in the other boarding systems are intended to prepare children for entry into the world of work. Only the best students are expected to continue in school. The boarding system has developed rapidly over the last decade—there were only forty-three boarders in 1980, but nearly 3,000 by 1990, representing 22% of the total number of Yi pupils in the whole county. About 25% were in the key-point boarding system (including some secondary school pupils), slightly more in general boarding schools and the rest in the semi-boarding system. Boarding pupils apparently achieve better as a result of having more time to learn and more consistent attendance. They are disproportionately successful in gaining access to higher levels of education. Most of the cost of the boarding system is provided by subsidies from higher administrative levels, with only a small part granted by the county government.

The boarding system is popular with officials and teachers who believe that it is necessary for educational development. Many Yi parents cannot afford to send their children to school, or appear to be less than enthusiastic about the idea. Boarding schools help retain children in education at higher levels, and the system of streaming students into different types of boarding schools is defended by arguing that the state cannot afford to put all students in the high cost schools, and that it is appropriate to invest more in the best pupils.

Lastly, language policy is an area of special concern. Yi people have their own language unrelated to Chinese. The language of instruction in the first schools was Chinese. Most people do not understand Chinese as Yi is the basic language of communication. The Constitution and the 'Act of Autonomy in Minority Regions of the People's Republic of China' states that 'the autonomous region's government has a right to act on its own to develop education for minority people,' and to 'select for use in instruction languages from the region'. Some of the schools in rural areas started to offer Yi in 1979, and the primary and secondary schools in the county town began offering Yi for boarding classes in 1982.

Since 1986, two systems of bilingual instruction have been practised. System 1 uses Yi as the medium of instruction, and Chinese is offered as a subject from grade 2 to grade 6. The pupils are expected to master more content from Yi textbooks than from the Chinese ones. System 2 uses Chinese as the medium of instruction and Yi is offered as a subject from grade 3 to grade 6. Here, the emphasis is on the content of Chinese textbooks. There is some local variation of the two systems but the basic pattern is retained. In 1991, there were five xiang and two county town schools piloting system 1, with 1,200 primary school pupils involved. System 2 was used in twenty primary schools with 1,300 pupils. About 23% of the pupils are in either system 1 or 2 schools. Chinese is the only medium of instruction for the rest of the pupils, though teachers still use Yi to explain what the children do not understand in Chinese. In 1990, the first 112 pupils graduated from the primary schools of System 1 and eighty were admitted to Yi medium lower secondary schools.

The bilingual programmes have a very short history, and problems of teaching quality, and lack of availability of written materials, are still serious. Many people continue to have doubts about the bilingual programmes, especially regarding System 1. The main criticisms made by those who were interviewed were that:

— since there are no schools at higher levels for the graduates from the schools of System 1 further education opportunities are limited;

— the examinations held for recruiting government employees or workers are all in Chinese, so the employment chances of Yi stream students are restricted;

— time spent learning Yi reduces time spent learning other subjects; and

— the Yi language is too simple to explain the complex modern world, there are problems in translation, and it is easier to teach subjects like science in Chinese rather than in Yi.

In contrast to these possible disadvantages are the positive effects that making more use of the mother tongue may have, such as raising levels of achievement, motivation and participation. Nevertheless, judgements on the educational consequences of the bilingual policy in Zhaojue require more data than is currently available. Since similar debates take place in relation to other minority groups (e.g. Zhang Yuanqing, 1985; Wang Lianfang, 1985) it would seem important to collect systematic data on this before reaching any hard and fast conclusions.

Discussion

This case study raises a number of issues that are common to the provision of education to minorities in other parts of China (Wei & Zhou, 1984). The data presented here show that despite the various efforts that have been made, enrolment and school completion rates remain very low. The rates for girls are strikingly small, especially given what is known about the direct and indirect benefits of investing in the education of females (Lewin, 1994, p. 20). The causes of poor performance on basic indicators of participation are multiple and complex. Amongst the Yi in this area they include:

— the traditional valuing of boys over girls and various cultural practices (arranged marriages in childhood, dowries, sibling care) encourage girls to drop out or not to enrol at all;

— the lack of sufficient primary school places for all children reflecting the fact that educational provision has not been regarded as a high priority in practice;

— the low quality of physical facilities and the very limited availability of educational materials which impoverish the learning environment in many schools;

— irregular attendance of both pupils and teachers, especially related to the agricultural cycle;

— the lack of quality and motivation of teachers;

— low levels of educational achievement leading to repetition, drop out and push out; and

— the problems that arise from low levels of parental literacy and the use of two languages in the schools.

Attempts are being made to address the language issue through the two systems which have been introduced. The debate on language medium is a familiar one, and has been running for a long time (Watson, 1981, p. 111). There is a predictable tension between the national language (Chinese) and the mother tongue (Yi). The latter will never have much value outside the Yi community and, though Yi language is central to Yi culture, the development of the economy is likely to depend on those who develop links with enterprises in the rest of China. These will be undertaken in Chinese. Higher levels of schooling are likely to continue using Chinese as the medium of instruction. The language question, therefore, concerns when the actual transition to Chinese should take place, and for how long it is viable to teach using Yi as the medium of instruction. It also concerns the relative costs and effectiveness of different approaches to bilingualism. Currently it is not clear what these might be, although all indications show that systems which begin the transition at a later point in the education cycle are likely to be more costly.

The interventions that are being introduced to encourage sustained female enrolment raise a number of questions. Although these programmes have been in place for some time, female enrolment remains very low. The girls who benefit from the subsidies and incentives may also include many who would be in school in any event—the girls in city schools come from backgrounds where parents are more likely to send them to school. The question is, which of the measures—girl only classes, boarding, higher allowances per child—actually make a significant difference to retention and performance. And what strategies would encourage more rural girls to attend? The fact that attendance remains so low suggests that root causes may lie outside the school system rather more than inside it. Tradition and some aspects of economic reality appear to mitigate against female enrolment. It is not clear that community level approaches are being mobilized to influence parents' attitudes. There is a dilemma here, since Yi culture would be completely transformed by more universal access to schooling for girls. Nevertheless, the Yi community has yet to embrace the idea and until it is ready to do so, change has to be in some sense imposed from the outside.

Sanctions are employed to encourage enrolment in general, and punitive measures have been taken against some parents. This approach may be bureaucratically attractive but is flawed for several reasons. First of all, it is not easy to trace large numbers of non attendees. Even if most could be identified, all could not be enrolled in the short term. Also, fines may not be appropriate for families with little or no cash income and few assets. It seems, however, that parents are gradually becoming aware of their legal responsibilities to enrol children in school. Whether they do or not will depend more on the availability of places at school, the quality of the facilities, and the perceived value of the education offered by schools, rather than on legal sanctions that are difficult to enforce.

The relevance of schooling is perceived by many teachers as problematic. Half the teachers we questioned at Bier and Sikai central primary schools attributed lack of 'learning motivation' as the primary reason for low achievement and subsequent drop-out. This in turn they saw as related to the lack of opportunities to benefit from schooling through access to non-agricultural jobs and higher levels of education. There are very few opportunities for educated Yi students to aspire to higher studies, and very few successful role models, especially for girls. This is neither a problem that can be solved easily, nor is it restricted to the area of this case study. Cleverley (1990, p. 6) notes that this is a factor in rural education in many parts of China. However in the Yi area, positive discrimination is practised to create more opportunities than would otherwise be the case. Quotas are employed and entry requirements lowered to allow more pupils through to higher levels. The question is, should more concessions be made, or would doing so merely make it more likely that those who progress to higher levels will subsequently drop out? It would seem that any further lowering of the standard for primary school graduation is an unappealing alternative.

Low achievement levels may result in part from the short length of the school day, and from high rates of absenteeism. Transport is poor and many pupils live very far away from their schools. Classes may not commence until late morning (11:30), and end by mid-afternoon (3:30). Pupils in village schools may be in school for even less time, as many of these schools offer only two or three class periods a day. Irregular attendance is widespread and seriously disruptive of learning continuity, yet curriculum materials are designed as though attendance were continuous, and the question here is whether it is possible to design curriculum materials that are more modular and free-standing so that irregular attendees can engage meaningfully with learning that does not assume continuity of previous experience.

The great majority of teachers in Zhaojue are qualified if we judge them according to their educational background. Most are graduates of Zhaojue Normal School and other normal schools in Liangshan Prefecture. These schools have special entrance quotas and levels of achievement are low. One third of the graduates are judged unable to teach according to county officials we interviewed, and one third require additional in-service training before they are allowed to teach. What may be even more serious is that many teachers want to transfer to other fields, and as a result may not feel much commitment to their work. Teachers were asked to respond to a question asking if they would rather have another job—71% of the teachers in the Sikai sample and 50% in Bier indicated their preference was for a job outside that of the teaching profession. The average yearly income of teachers is about ten times the net average income per capita of the rural areas in Zhaojue; in spite of this, many still want to change their jobs. The reasons they gave for this included lack of respect from the communities they serve, lack of opportunities to earn additional income, the difficulty of achieving urban residence and physical insecurity.

The actual emphasis on education in development policy depends on the quality and commitment of the educational administration. The officials we interviewed at each administrative level argued that they placed great stress on education and gave it first priority in line with national directives. Where it proves difficult to translate this ambition into practice, inadequate financial support is frequently identified as the main difficulty. In contrast, the staff of educational institutions we spoke with tended to argue that government officials 'only stress education orally but not in their actions, and they emphasize education in meetings but not after meetings'. It appears that many government officials do not want to take responsibility for education. A typical view is that 'family planning is the most difficult work to do and education is the most complicated. It is not easy to see accomplishments in education and gain recognition'. The question is, what is the level of local commitment to the development of education and are the resources made available sufficient to allow real progress towards defined goals?

Marginality may come in several forms. The Yi themselves are marginal to mainstream Chinese development, and their educational needs have been neglected in the past for this reason. Within the Yi area, core-periphery relationships exist between the towns, villages and the rural hinterland. Those who administer and invest public resources may place education high on their agenda. If this is so, it is difficult to understand why there are so many areas without reasonable access to schools and—where there are schools—to find classrooms with no furniture. Equipping such schools would require an investment which is much less than that made available for some other local development projects. The point is that the development of education amongst the Yi may depend more on the preferences and interests of relatively local groups than it does on national level support and subsidy, and national policy. Bureaucratic responses to poor system performance need to be tempered by the realization that out-of-school factors may be at least as important as in-school factors in determining the outcomes of equity in education.

Note

1. This paper is based on work reported in Lewin, K., Wang Y., Qu H., Wang L., Li J., Wu Z. and Quian J., Implementing basic education in China: progress and prospects in rich, poor and national minority areas, IIEP, 1994. I am indebted to the members of the team from Beijing Normal University who made this research possible. The research was supported by UNICEF and the British Council.

References

Ahmed, M, et al. 1991. Basic education and national development: lessons from China and India. New York, UNICEF.

Cleverley, J., ed. 1990. Rural education: policy and programmes. Chinese education (New York), vol. 22, no. 4, p. 3-8, Winter 1989-90.

Kwong, J., ed. 1989. Education of minorities. Chinese education (New York), vol. 22, no. 1, p. 3-7, Spring 1989. (Special issue.)

Lewin, K.M. 1994. Education and development: the options and the evidence. London, British Overseas Development Administration. (Research monograph.)

Lewin, K., et al. 1994a. Educational innovation in China: tracing the impact of the 1985 reforms. London, Longman.

Lewin, K., et al. 1994b. Implementing basic education in China: progress and prospects in rich, poor and national minority areas. Paris, UNESCO/International Institute for Educational Planning. 174 p. (IIEP research report, 101)

Wang Lianfang. 1985. The problem of developing minority education in Yunnan province. Reprinted in Kwong, op cit., p. 21-26.

Wei Shiyuan; Zhou Guangda. 1984. Development of education programs in mountainous regions to enhance the culture and knowledge of minority nationalities. Reprinted in Kwong, op cit., p. 27-46.

Watson, K. 1981. Changing policies towards the education of minority groups in China. In: Education of minorities world yearbook of education 1981, p. 98-115. London, Kogan Page; New York, Nichols.

Zhang Yuanqing. 1985. Current problems of education and proposals for improvement in Tai nationality schools in Gengma, Tai-Wa Nationalities Autonomous County, Yunnan. Reprinted in Kwong, op cit., p. 47-55.