Clouds blot out the warm Autumn sunshine and a light dusting of early snow swirls up the valley as Mr. Laba Sornam begins to give a grade four Tibetan class in Kalowa Primary School, some 90 kilometres north east of Lhasa.

More than half of the 276 children at this school board in dormitories behind the classrooms for eleven days at a time. They trek at altitudes of 4000 metres above sea level for up to seven hours from their homes in outlying villages, carrying sacks of tsampa - ground barley that is mixed with sugar and tea into a coarse paste - that will be their daily bread for the extended school week. Some from better-off homes also carry yak butter to mix with their tea. After eleven days school, they trek home again for a three day rest.

Classrooms and dormitories are unheated, even in winter when temperatures can drop as low as 20° Celsius. In compliance with national regulations, a basketball court, running track and soccer pitch are marked out in front of the school; but there are no balls and, even if there were, the pitches are so covered with rocks that nothing resembling the intended game could be played there.

The school has 13 teachers, eight classrooms with blackboards and desks, and almost nothing else by way of teaching reources. Although in theory the curriculum includes social studies, science and ideological education (sixiang pinde), in practice these are rarely taught outside of cities and some county seats, because the children cannot afford the necessary textbooks and the teachers often lack the knowledge and skills to teach the subjects. In Kalowa only three core subjects are regularly taught: Tibetan, Chinese and Math. The children do not draw, because there is no paper to draw on. There is no music.

This, in short, is a fairly typical rural Tibetan primary school. Some villages have smaller, and even more rudimentary, day schools that teach only grades one to three. These usually have only one or two teachers, who often have to teach two grades at once, slipping between class-rooms and leaving children to work unsupervised. On graduation, the children must switch to a boarding school to continue their education. Many don't.

But if Kalowa school is typical of its kind, Laba Sornams Tibetan class is unusual in its approach. He begins by pinning to the blackboard a hand-drawn picture of a solar oven, and asks for volunteers to say what it is and what it is used for. He then writes the Tibetan script for solar oven next to the picture. (Tibetan script, like classical Chinese, differs from the spoken language, making new written vocabulary harder to absorb.) Next he shows the class flash-cards for six new words they will study today. The children are asked to read a short passage from their textbooks and spot where the new words occur. Then they work in groups of four to discuss and suggest definitions for the words, writing the answers on slips of precious paper. (The desks are arranged in groups of four, rather than the customary rows, to facilitate group work of this kind.) The groups report back on one word each, and the others are asked if they agree on the suggested definition.

These attempts to actively engage children in the learning process represent, for all their modesty, a profound departure from normal Tibetan, and indeed Chinese, teaching methods in rural schools. Children are more usually treated as empty vessels: passive recipients of knowledge that is transmitted only via the teacher and the textbook. The teacher speaks, reinforcing the set texts, and the children listen. Examinations are geared almost exclusively to determining how much of the set texts the students have remembered.

That children can help each other learn; that investigation, experimentation and articulating their own ideas are more powerful learning experiences for children than merely listening; that it is appropriate to develop critical and analytical faculties as well as rote learning; that cheap, home-made teaching aids can enrich the slender set texts . . . These all remain largely foreign ideas. But, as the Chinese government embarks on a search for quality education (see this article:Exam system is the weakest link in quality chain), these are ideas that officials and educationalists are increasingly willing to explore.

Laba Sornam and his colleagues in Kalowa Primary School are exploring the ideas in practice as part of a Save the Children (UK) basic education project carried out in partnership with the Lhasa Municipality Education Bureau, with funding support from the British government. As part of the project, teachers from Kalowa and other schools in two townships of Medrogonkar County, Lhasa Municipaility, have each attended a total of 16 days training in teaching methodologies, focussing on four key areas: questioning; group work; introducing games into the classroom and using real, locally available objects as teaching aids.

The questioning component of the training encourages teachers to put questions to the class that are open in two senses: any child can offer to answer them (rather than one child being put on the spot); and the questions are designed to make children think, or express an opinion, rather than parrot a right answer. Mr. Pasang, another Kalowa teacher, puts this in context:

There are two main reasons why children don't come to school, he says. One is money, of course, if the children cant afford the books or uniform. But another is that some teachers point at a child, and make them stand up to answer a question. If they get the wrong answer, they may be made to stand up for the whole lesson. So some children hate school. This is why we must use child centred teaching methods.

Local education officials are also swayed by marked improvement in exam results from the first pilot schools. This encouraged them to allow the project to be scaled up to other schools.

The concept of group work is now well accepted by the teachers. At first it was quite difficult to get children to discuss and work in groups, but now they do it automatically, says Laba Sornam. His colleague, Mi Ma, adds that group work is particularly useful in Chinese classes because the children find Chinese very hard, and this way they really help each other.

Games are also emphasised in the trainings as a cheap and easy way to enliven learning. Trainees are shown how cardboard from used cartons can be made into durable Bingo cards. The cards are divided into squares, in which math equations (16 + 8 = ?), or Chinese characters are written. The teacher holds up a flashcard with a number or a word in pinyin, and children check the boxes on their cards. A variation is finding a friend: children are each given one card an incomplete equation or character and must find the classmate with the corresponding card. In an even more chaotic game, used for teaching the time, children are organised into a human clock: a small phalanx of human hour and minute hands moves around a circle of classmates holding numbered placards, to point out different times of day. A variation is to create a human abacus, with phalanxes of units, tens, hundreds and thousands.

Imaginative use of low- or no-cost, locally available, real objects as teaching aids is another focus of training. Sticks and stones for counting; a potato for cutting up into fractions; a chair for illustrating angles and dimensions. This kind of appropriate educational technology, says Save the Children Education Adviser, David Strawbridge, is not only a cheap substitute for more expensive classroom aids but also helps to root formal education in the real, familiar world that the children inhabit. Moreover, it engages all five senses in the learning process - not just the ears and the eyes - reflecting the way that young children learn in the home environment

Over the ten years that it has worked in basic education in the Tibet Autonomous Region, Save the Children's programme has steadily evolved towards delivery of this kind of educational software. In the early 90s, the organisations entry point was primary school construction in remote, rural areas. From the mid 90s, the programme began to support in-service teacher education, delivered through local training institutions, to improve the knowledge base of village teachers, many of whom had themselves only completed junior high school education.

At the same time, the organisation began a series of small-scale income generation projects for rural schools. Around a dozen grants ranging from USD 120 to USD 400 enabled schools to develop vegetable gardens, greenhouses, village shops, and yak breeding projects. According to an external appraisal conducted last year, most of the projects showed a return of around USD 300 per year in cash or kind, and this was used for supplementing children's diet in boarding schools; buying candles, bedding, stationery and teaching aids; providing modest prizes for students and organising special events such as sports competitions. In the county town of Medrogonkar, a later grant has funded the opening of a restaurant that generates funds for local schools.

It was in 1998 that Save the Children first ventured into work on teaching methods in the TAR, with a workshop for primary teachers in Medgrogongkar's Tashigang township, facilitated by an overseas trainer, to introduce the concept of child-centred education. This initiative was partly in response to the regional authorities change of track on in-service teacher training (discussed below). It was possible because enough trust had been established with local partners for the organisation to promote new approaches especially since these now resonated with policy pronouncements coming from Beijing.

More detailed training sessions followed, concentrating on the selected four key areas, in the townships of Tashigang and Tsashol. The project has since spread to two townships of neighbouring Lhundrup County.

These initial workshops and follow-up, refresher sessions, which include demonstration lessons, were facilitated by Save the Children's Chinese and expatriate staff, who spent up to ten days at a time in the field, sleeping in the school dormitories.

In order to develop a support system for staff in the smaller, outlying day schools, key teachers from the boarding schools are appointed to visit and mentor their more isolated colleagues, sitting in on classes and commenting on the methods. Some of the key teachers participated in a study tour to Nepal's Solokumbu Province, where a similar programme of outreach supervision has been sponsored by the Himalaya Trust.

David Strawbridge hopes that from the key teachers will emerge a corps of skilled communicators capable of assuming a training role to extend the project to other areas. Promising candidates have already been identified.

This is an unusual experiment in building grass-roots capacity that can then spread horizontally, rather than relying on a vertical, training of trainers model. Cascade or T-o-T training approaches are more frequently seen by international organisations as the answer to China's daunting problems of scale: the only effective way to reach large numbers. But this can be a problematic approach in the field of education, for the simple reason that China's top educationalists often have little or no experience of teaching children. Professors in the Normal Universities are chosen for their academic standing, not for their classroom experience, and pre-service training focuses almost exclusively on the content of subject areas to be taught in schools, with only cursory attention to teaching methods. As a corollary of this, good teachers are generally regarded as those who are knowledgeable in their subject area, rather than those skilled in communicating their knowledge or stimulating children to learn for themselves.

The World Wide Fund for Nature China Programme has wrestled with this problem in a rather different context.

The Fund's Environmental Educators Initiative, supported by BP-Amoco and NOVIB, has worked for six years with the Ministry of Education, the Peoples Educational Press and prestige Normal Universities in Beijing, Shanghai and Chengdu, to develop and pilot primary and middle school environmental education materials and curricula. Several other international organisations that have sponsored environmental education programmes in China (e.g. Wildlife Conservation Society, US-China Environment Fund, and the Global Environmental Alliance) have found that these are often popular with Chinese schoolteachers not only, or even mainly, for the environmental content, but also for the more active learning approaches that the programmes generally involve. WWF's is no exception. It aims, no less, to foster critical thinking about the environment and related social issues, and to encourage students to become active, environmental citizens. As well as engaging in field studies and surveys, children in pilot schools have gone out into their communities to take direct action such as lobbying restaurants to give up using disposable chopsticks and persuading market vendors (and their customers) to use fewer plastic bags.

In addition to training key teachers in pilot schools and producing textbooks with lesson plans, WWF and the Normal University professors who were initially trained in the programme have worked closely with the Ministry of Education to develop a new, national curriculum for primary and middle school environmental education. (In order to avoid over-burdening the existing syllabus with a new subject area, the curriculum cuts across all subject areas as in any case befits an ecological perspective recommending ways to introduce environmental awareness into everything from Chinese literature to Physics.)

This remarkable achievement has been the result of a long and painstaking process. It began with the formation of Environmental Education Centres at the three Normal Universities, drawing together lecturers from different subject areas. These master teachers attended a series of training workshops facilitated by international environmental education educators, and some enrolled in relevant Masters degree courses at universities overseas. The master teachers then worked with teachers in pilot schools, who also attended workshops with overseas trainers, to design and test classroom activities across different subject areas. These experiences were reviewed in joint workshops where staff from the three Normal University centres worked together on developing and revising teaching materials.

This approach appears to have been successful, and the project is now being scaled up or, rather, outwards to embrace teacher training institutions in other cities. One advantage of starting with the elite Normal Universities has been that these are also important think tanks in curriculum development for the Ministry of Education. This, and close working relations with the curriculum reform department of the central Ministry itself, have made it possible for the project experience to feed directly into the development of the new, national curriculum.

However, one of the main challenges in the programme has been the lack of classroom experience of the master teachers. They have a lot of theory, but not a lot practice notes Programme Coordinator, Ms. Liu Yunhua. This naturally reduces their insight into class management, and so adds to difficulties of educating the educators in new ways of engaging and stimulating children.

These difficulties are likely to be greater when dealing not with relatively privileged and well-equipped metropolitan universities and urban schools, but with poor, rural schools. It is likely to prove harder to find a common language between university professors and rural school-teachers; and, as rural Tibet amply demonstrates, policy or curriculum developments at the top do not necessarily trickle down to classrooms in outlying areas. These are issues WWF will soon come up against, as Lhasa Normal School is one of the institutions participating in the expansion phase of the Environmental Educators Initiative.

Having started from the chalk-face in the poorest schools, the challenge for Save the Children will, by contrast, be how to extend and institutionalise the model it has demonstrated. A second phase of the project, which the organisation hopes to begin from next year, will aim to spread the work to Nyimu, and Dhamshong Counties in Lhasa Municipality, and to Rombhuk County in Shigatse Prefecture, as well as working with the Lhasa Normal School to support in-service and pre-service training. For the present, however, efforts are concentrated on making the model work, and demonstrating that it really is possible to achieve better classroom methods within existing financial and human resource constraints.

November will see the first of a series of workshops in multi-grade teaching methods for teachers in the smaller schools, to help them deal with the demands of teaching more than one class at a time.

Save the Children staff are also accompanying key teachers on mentoring visits to the smaller, village schools, to provide hands-on support in developing this supervision approach.

At the same time, the organisation is trying to integrate other areas of its work in the TAR into the school-based project. An ongoing community water supply and hygiene education programme that has served some of the same communities is now delivering hygiene education training to schools in the pilot townships; and attention is being given to creating synergies between schools and out-of- school community literacy activities.

Save the Children has already worked for several years with the education authorities to strengthen their delivery of adult literacy classes. In the view of programme consultant, Udaya Manandhar, these have in the past tended to be driven by quotas, with each county given a yearly target of illiterate adults to educate through formal classes. In such a system, he says, literacy is too narrowly measured by proficiency in a test that is very closely linked to the curriculum, rather than to real life situations. Moreover, there are few post-course materials for graduates to read, or contexts for them to apply their newly acquired skills. As a result, many people soon forget what they have learned. To counter this, Save the Children is promoting a community literacy approach that seeks to build on the existing curriculum by developing and including materials relevant to vocational skills and income generation, with practical applications in helping to improve livelihoods. Involvement of parents in the management of the school based income generation projects is one way of furthering this goal, while also contributing to another overarching objective: building closer parent-teacher relationships, and achieving greater parental involvement in children's education.

Approaches developed in Medrogongkar and Lhundrup will be replicated not only in the anticipated, second phase, but also in Shigatse's Panam County, where Save the Children is implementing the education component of a long-delayed, European Union funded integrated development project (see this article:Abandoning the farm for greener fields of governance). In Yunnan, too, Save the Children is working on the same issues in a project that, since 1999, has aimed to improve the reach and quality of basic education in poor minority areas of Simao, Lincang and Dali Prefectures.

Meanwhile, the US based Trace Foundation is also designing a project to enhance the quality of basic education in the TAR. This is expected to begin early next year, probably initially based in selected counties of Nagchu and Lhoka Prefectures. It too will comprise in-service training with special attention to teaching methods and the professional development of teachers, a mentoring system involving core teachers, and the production and dissemination of teaching materials. Whereas the Save the Children project in Tibet has drawn from the experience of similar work in Nepal, the Trace programme will attempt to adapt some of the experiences of a School 2001 project carried out in Mongolia by the Mongolian Foundation for Open Society.

Tibet has much in common with other poor, minority areas, but there are also critical areas of difference. One such is educational financing. In some respects, money is less of a problem in impoverished Tibet than in what Tibetans call inner China.

County governments in other provinces have to cover their own educational expenditures, and in many cases are unable to do so, leaving teachers without pay for months on end. But in the TAR (where rural people are exempt from agricultural taxes, leaving local governments with negligible revenue) teacher's salaries are covered by subsidies from central government. What's more, salaries for public servants are much higher than in the rest of China: a qualified teacher can expect to start on CNY 1,000 - 1,200 per month (USD 120 - 146), more than twice the rate elsewhere. Pay scales within the TAR are linked to altitude, as a reflection of presumed hardship: the higher the work-place, the higher the wage.

Yet there is still a shortage of qualified teachers. Many rural schools remain significantly dependent on minban (roughly 'community') teachers who are the pedagogical equivalent of the barefoot doctor: local villagers with no more than junior high school education. In the TAR, these teachers earn CNY 100 - 200 per month (USD 12 - 24).

During the early years of the reform and opening period, in the drive to extend the reach of basic education without excessive demands on government budgets, many minban schools were established across China using the labour and resources of local communities. These were generally very modest buildings staffed by essentially unqualified minban teachers. At that time, almost any teacher seemed better than none.

From the 1990s, government policy was to professionalise the minban teachers through in-service training, until they could qualify as fully-fledged, state schoolteachers. To this end, local training institutions began to offer certification courses, generally through summer schools. As in pre-service training, the courses dealt with the content of the taught subjects, setting out to compensate for the high school education minban teachers never had.

Yet the poorest areas often lacked the resources to offer training, and teachers could seldom afford to attend. Many international agencies therefore provided support for in-service training, in the TAR and elsewhere.

By the late 1990s, some progress had been made in upgrading teachers, although official accounts of progress were often exaggerated. Some local authorities, receiving from on high quotas of minban teachers to upgrade, responded by creating a new category of unqualified, low paid, daike ('substitute') teachers on a temporary basis, until sufficient numbers of fully qualified teachers, and the funds to pay them, could be found. This enabled local authorities to report a steep decline in the numbers of minban teachers, although actual achievements were more modest.

Central government, however, continues to pursue an ambitious vision of a post-2010 China in which all teachers will have completed high school or equivalent (12 years education) and a further two years of teacher education at tertiary level. For remaining, unqualified teachers, whether called minban or daike, the qualification threshold is moving further away.

In the TAR, in-service certification courses have been suspended and county and prefectural level training institutions have been closed down, leaving the Lhasa Normal School as the only teacher training institute in the Region. The school is being upgraded, from a zhongzhuan (roughly equivalent to vocational high school) to a dazhuan (roughly equivalent to college) institution.

But even so, it is unlikely in the short-medium term that Lhasa can ensure an adequate supply of fully qualified, local teachers. How will the deficit be made up? Some foresee a greater influx of teachers from inner China. Already, graduate teachers from eastern provinces are recruited for middle school posts in Lhasa and other main population centres. According to Mr. He Chunlin, head of the Human Resources Department of the Lhasa Municipal Education Bureau, there are around 300 teachers from other provinces working in Lhasa, and as many as 1,000 in the TAR as a whole.

Some of the shortfall in the teaching stock may be made up by Tibetans returning from full-time education elsewhere in China. Since the mid 1980s, Tibetan children with the highest exam scores at the end of grade 9 have been offered state scholarships to complete their high school and, if they qualify, tertiary studies in other provinces. Although this often involves family separation lasting many years, parents generally see this as presenting irresistible opportunities for the future advancement of their children. Critics of the scheme are concerned that it amounts to cultural uprooting of the children selected, and the training of a local elite that will more willingly embrace Han influence in the future. The government can respond that it is offering Tibetan children the best educational opportunities at its disposal, in the absence of an adequate training infrastructure in Tibet. Returnees from the east, it might add, are an essential human resource for building up local educational capacity. Once there are enough qualified Tibetan teachers, Tibetan children will no longer need to go east for their high school education.

But the drive to raise the qualification bar for schoolteachers will, if successful, also force less qualified, local people out of the profession. He Chunlin estimates that there are a total of 500 minban or daike teachers remaining in the seven counties that make up Lhasa municipality. Since 1997, he says, these teachers have had the opportunity to sit an entrance exam for a full-time, two-year course at the Lhasa Normal School. Graduates from the course will become fully qualified state teachers. Those who fail, after a maximum of three attempts, to gain entrance to the course, will remain in post only until such time as they can be replaced by fully qualified teachers. This year, according to Mr. Wang, is the cut-off year for applications. Remaining minban teachers perhaps one fifth to one quarter of all teachers in the TAR -- are thus on the short-list for removal.

Some in the smallest and most remote communities have already gone – their schools closed down or merged with larger, township facilities.

According to some reports, this policy has resulted in a significant drop in primary enrolment. Parents who sent their children to a local day school appear unwilling or unable to send them away to board. The expense of doing so is presumed to be a major deterrent. Save the Children has estimated the total cost to a family of placing a child in school, in terms of fees for textbooks, clothing and stationery, at CNY 46 per year for the first three years, thereafter rising steeply to CNY 274 per year. Although this is considerably less than in most areas of China, it is a substantial sum in Tibet, where the subsistence economy provides minimal opportunities for earning cash income, and where many families have three, four or more children. Financial barriers are generally held to account for low access to basic education in TAR, where only 30% of children are officially reported to complete 9 years.

In an apparent attempt to address this, the TAR Education Commission recently announced a new direct subsidy package for all children enrolled in schools more than 2 kilometres from their home: each child is to receive a grant of CNY 100 per year for uniform (in practice, at present only children in cities and county centres wear uniforms all year round; in rural schools, children dress in whatever their parents can put together for them, and those that can afford the track-suit style uniforms keep them for special occasions); and a further grant of no less than CNY 500 per year for materials and subsistence.

Schools are still in the process of identifying potential beneficiaries of the scheme, and details of how it will work in practice remain to be clarified. But it appears nonetheless to signal a major effort to get Tibetan children into school, and local education authorities report that there has already been a surge in enrolment as a result.

Government policy towards basic education in the TAR thus involves substantial investment in teacher pay, training and, now, direct subsidies to children. Clearly this depends upon a high level of financial transfusion from the centre. Given the relatively small size of Tibet's school age population (only 360,000 enrolled primary students) and the political importance to China of being seen to narrow growing east/west differentials (and, especially, to deal fairly with Tibet) this may be an affordable proposition. But it would cost a great deal more to extend to for other western provinces - Guizhou with its 5 million enrolled primary children, or Guangxi with 5.6 million.

Even in Tibet, the plans may be too grandiose, and too removed from the reality on the ground, to produce the desired results in the short-medium term. Schools such as that at Kalowa are struggling to deliver just the basics: literacy and numeracy. Is it likely that they can leap-frog to an altogether higher level of provision?

Policy circulars, landing from above on the desks of local education officials, are pressing for evolutionary leaps. (These do, however, commonly include a caveat to the effect that the policy should be implemented when and where conditions permit). This year, for example, has seen guidelines recommending that Math be taught in Chinese from grade one, and that English classes be introduced into the curriculum from grade four.

Kalowa school Principal, Mr. Chundra, appears bemused by these suggestions. He does not speak Chinese himself. Chinese language teachers at the school introduce pinyin from grade one, but progress is slow and achievements modest. Most of the children start school with no Chinese at all. Interviewing a couple of dozen students, we found only one who had ever been to Lhasa; many had never even travelled the 40 kilometres to the Medrogonkar county seat. They therefore have negligible previous contact with Chinese language, and negligible opportunities for practice and reinforcement of what they learn in class. As to English: although it is spoken by some better-off, urban Tibetans who were educated in India, none of the Kalowa teachers, let alone the students, can say so much as Hello.

It is only in the last few years that any Chinese has been taught in rural Tibetan schools: previously, they concentrated only on Tibetan and Math. The addition of Chinese does in theory represent a broadening of opportunities. According to Save the Children Programme Officer, Ms. Pelky, When we did our base-line survey, we found that parents really wanted their children to learn Chinese. They see this is as a way to improve their chances of doing business or getting a job. Proficiency in Chinese is also essential for children to go into middle school and beyond, where it is the main language of instruction.

There is no doubt, however, that the incorporation of Chinese and other subjects into the curriculum make it harder, by reducing the amount of time devoted to Tibetan, to ensure that children will at least achieve basic literacy in their native tongue. The policy therefore worries those who fear the steady erosion and marginalisation of Tibetan culture.

Proficiency in several languages is not, of course, beyond the natural capacities of human beings; but it has to be asked whether, in setting out these aspirations, the TAR Education Commission is not expecting the Region, its children and teachers, to run before they have managed to walk. Given the situation in rural areas, would it not be more realistic to concentrate first on making the few years that most Tibetan children spend in school as productive, useful and relevant as possible?

The Save the Children project has set out to show that it is possible, without great financial investment, to make classrooms more inclusive and active centres of learning by changing the nature of the relationship between teachers and students. This can have immediate impacts even in schools that remain poor, with limited human resources.

If the education authorities succeed in expanding the curriculum and the proportion of fully qualified teachers, it will almost certainly still be necessary to re-examine the implicit assumption that more education and more educated teachers naturally equate with better quality of education. Thought will also have to be given to the methods used to deliver education in the classroom. This is likely to entail debate about what education is, and about what it is for.

It is at least possible, however, that the years to come will see a growing gulf between educational standards and practice in urban and rural areas of Tibet. Bilingual education, for example, is relatively easy to achieve in the towns, where already kindergartens (unknown in rural areas) often use both Tibetan and Chinese. Urban primary and middle schools usually offer the full, national curriculum, with a level of provision that is broadly comparable to inner Chinese provinces. But there are few signs that rural schools are keeping pace; and the wider the gulf that opens up, the harder it will be for rural children to cross it and compete in the - invariably urban - middle schools.

Hence the importance of starting from the bottom and making the most of what is available to improve learning experiences in the rural schools. This, of course is relevant not just in Tibet but in the many poor areas of China that do not enjoy the same subsidies from central government.