SAIDE's Study of Education Project is developing nine mixed-media modules in education studies for use in pre-service and in-service education of teachers. Aimed at superseding the fundamental pedagogics texts available for teacher education in South Africa, the Becoming Competent series attempts to model a form of resource-based, distance education courseware that can also be used in traditional face-to-face teacher education.

Each module team consists of a coordinator and from two to five other core writers. In some cases between one and three additional 'sub-writers' are commissioned to write short pieces such as case studies. All of these are contracted to write in a part-time capacity, the only full-time personnel being the Project Leader and the Project Manager/Assistant Leader. Altogether, almost forty writers have been, or are, involved. The module coordinators and other writers meet and write mostly in their own time while holding lecturing or other posts. Three of the writers are freelance materials developers. The membership of most teams was, for a number of reasons, geographically spread: for instance, in Durban, Johannesburg and Giyani; or in Pietermaritzburg, Durban and East London.

In order to meet the project's secondary goal of capacity-building, the composition of each writing team was very mixed with regard to writing skills, academic expertise, and institutional culture (mainly universities, colleges of education, and NGOs). However, it is worth noting that many of the more skilled academics struggled as much as the less experienced writers to come to terms with the particular difficulties of writing interactive materials for resource-based learning. One university professor claimed that he had found it less taxing to write six academic papers in one year (some published in international journals) than to co-produce one such module.

Our experience of collaborative (team) materials development has clearly been coloured by the project's developmental brief. This has probably resulted in a less-than-ideal collaborative model: several of the modules could in theory have been completed more quickly, and at less cost, if we had opted to employ only the most experienced writers in the field. However, the reality is that South Africa does not have an unlimited pool of academics and other writers who are, in addition, experienced and skilled in writing engaging, interactive, learner-centred materials. Hence (while the Project is explicitly aimed at increasing the experience pool referred to above) our experience is probably not far from the norm of current practice. As one might expect, there have been both benefits and frustrations in the collaborative (and developmental) process we have adopted.

Of interest perhaps to those considering a developmental agenda for contributors is that a number of inexperienced writers have shown clear gains in various areas (though it must be acknowledged that a few have opted out of the process within the first month or two of active writing). In one instance, a writer whose brief contribution did not eventually appear in the published learning guide, nevertheless openly expressed her appreciation of how much she had learnt, particularly about curriculum development, through her involvement in the development process (planning workshops and writing drafts). Another inexperienced writer who contributed to one of the first modules is now writing for one of the other module teams.

An obvious advantage of collaborative writing is the degree of representivity which the project has managed to achieve in composition of its writing teams, enabling a number of diverse voices and experience bases to have an impact on the content and form of materials: teacher educators from colleges of education, universities and NGOs, freelance materials developers, media producers; women and men; black and white; and so on.

Such synergic benefits are difficult to quantify. They feed into team discussion before anything gets written, and, through expanded perspectives on the part of writers, inform countless choices of content, emphasis, structure, vocabulary and style. An indispensable strategy for writers of learning materials is constantly to have one or more imaginary 'typical students' in mind to 'vet' whatever they are writing.

Working on one's own, or as part of a homogeneous team, it is all too easy to forget the differences possible in such an imaginary 'audience', and hence to write with a limited audience in mind. This is far less likely to happen if one is in frequent conversation within a diverse writing team, that includes writers who are actually teaching at the time.

Likewise, if there is a mix of regional contexts and of educational sectors (university and college, distance and face-to-face, secondary school and primary) represented in the experience base of the writers, there are likely to be gains in the local relevance of examples, illustrations, case studies and activities, as well as in the balance of theory and practice written into the text. Finally, an advantage for the Study of Education project was that writers could do some developmental testing of work-in-progress (such as we were able to do at all in such a short period) in colleges as well as universities.

It is probably fair to say of the process adopted for this project that the major difficulties presented by the collaborative approach related to the time it took most teams to complete their work - largely a consequence of employing part-time writers. If a team falls several months behind schedule, for whatever reason, the delay gives rise to a range of financial and administrative problems for both the project leadership team and the publishers.

Writers themselves who are held up may become frustrated by fellow team members, including coordinators, who fail to complete tasks according to schedule. In addition, writers who want to use published texts in courses at their own institutions find that prescription can be delayed by up to a year as a consequence of such delays.

Of course, this project's developmental aims precluded the use of only one or two 'experts'. Putting this consideration aside, it might have been easier and quicker to work with only one or two experienced writers (we were informed of one project, which, through experience, limited teams to three members). However, it would be wrong to conclude that writers working alone or in pairs necessarily complete work more quickly or efficiently than teams of, say, four to seven writers. This is because, depending on the degree of careful planning and on how writing tasks are divided amongst team members, it is usually quite possible for a number of writers to work simultaneously on different sections, even if they have to keep in mind that later adjustments will need to be made for the sake of coherence, unity and flow across the module as a whole. Neither approach guarantees to cut development time on the one hand, even one or two dedicated and experienced writers cannot write several chapters at once; on the other, a larger team is only as quick as its slowest writer, and it can be paralysed by indecision or inertia on the part of the coordinator. Even prompt delivery of a final manuscript does not necessarily imply that all is well - academic editing may, even at this late stage, highlight serious deficiencies in content or structure. In this event, quality goals may require substantial revisions, a process which may extend the final development stage by several months.

Project planners should take care not to underestimate how long it takes to write - and polish - quality learning materials. Here we are talking about materials that are engaging, learner-centred, interactive, issues-based, competence-led, inductive (taking the learner from particulars, and from the familiar and concrete, towards more theoretical and generalized understandings), and so on. The British Open University, after all, may allocate teams of full-time writers up to three years to develop modules.

While such ideal (and expensive) conditions will for some time be beyond the reach of most educational institutions in South Africa, it is probably fair to say that the eight months we initially allowed (from starting to identify the team, to submission of the final manuscript to the publisher) was too little - a year would have been more realistic. To allow much more than a year for developing a module of similar dimensions would be to invite writers - or more particularly coordinators, in our experience - to fall prey to Parkinson's Law (work spreads to fill the time allowed for it). Indeed, during major periods of delay, teams may be dormant while writers wait for the next cue from coordinators.The explanations for this are not especially surprising. Writers and coordinators who are chosen for their particular expertise or prominence in a particular field are all the more likely to be extremely busy people who are in great demand (and often away from their home base for periods of time).

For writers as well as for learners, the most difficult part of a demanding task is getting started. Workshops tend to waste time and other resources if a proposal has not been prepared by the coordinator, if key readings have not been done by all members, if all members have not written what they had been briefed to produce, or if they have not read one another's work.

If team members are separated geographically, this will impact on their ability to keep to schedule. Both writers and coordinators may constantly fall behind schedule if they are only infrequently face-to-face with the rest of the team (and the project leader). Without the prompt of others' proximity, it is all too easy to prioritize other tasks that are easier to get down to than writing. The corollary is that frequent proximity can allow for the discussion of points as they come up. The only team that was based in a single institution (since it was specially commissioned and had to be produced within a very short time), undoubtedly benefited - and delivered on schedule - because it was able to discuss small and large questions both informally and in scheduled meetings.

Most of the following points are based on our own experience, but a few of the strategies were contributed by members of the Project's Advisory Committee, and have yet to be implemented.

We learnt not to be over-ambitious: to provide more guidance and structure, and to add 40% to the time we had budgeted for writing. To these ends, the project leadership team modified and added to the guideline documents already provided. These included writers' guidelines, coordinators' guidelines (including a comprehensive schedule), and proposal guidelines. The detailed schedule listing every task, and negotiated with team members, was included in both writers' and coordinators' contracts - however, this was unfortunately no guarantee of prompt delivery.

While one may hesitate to include a penalty clause in the contracts of writers who are, after all, working part-time on the project while continuing to do full-time jobs (especially those who are prominent academics and are therefore likely to have many calls made on their time and energies), in the end there are de facto penalties that accrue if work is submitted very late, or in a far-from-finished state. Such 'penalties' include having one's work overwritten without time for consultation, or even not having one's work published at all. It is advisable to point this out to coordinators sooner rather than later.

A strategy which emerged late in the process, as a result of reflection and discussions with people from various institutions involved in team writing, was to extend the duration of the initial workshop (or possibly the second) from two days, to five days or even longer. Such a session provides an adequate opportunity for a high degree of clarity to be gained by the entire writing team, and for everyone to do a substantial piece of writing and revision of that writing. It also affords writers a better opportunity both to move past the 'getting started' hurdle and to experience, within a team, what it is like to produce quality inductive, interactive writing. The importance of fellow writers being in physical proximity has already been mentioned.

In partial contradiction to the foregoing strategy, Paul Musker, a member of our Advisory Committee with experience of collaborative writing in ELTIC, informed us that requiring a team to do early conceptualization at a long initial workshop can bog the process down and waste time. He advocated devoting longer joint working sessions (of a week or so) to resolving particular problems as they arose later in the course of writing. Paul also recommended supplying every team member with a selected pack of necessary common readings at the outset, rather than leaving them to their own devices.

The use of electronic mail to elicit and communicate comments on ideas, drafts and bits of drafts works very well to move the development process along - when writers make use of it. Unfortunately, the majority do not - even coordinators tended to use it mainly for administrative arrangements. Email discussion has enormous potential to offset the effects of geographic separation, but it takes extra time and effort over and above writing of the actual text. For this reason, 'small group' meetings of two or three team members in the same institution or city were encouraged in order to supplement more expensive team workshops.

A regular informative email 'newsletter' to all coordinators and writers helps to maintain momentum, especially if it mentions due dates for various drafts as a subtle means of maintaining pressure.

Mention must also be made of a factor that, while it may impact directly on the time it takes teams to deliver, has important implications in other areas as well. This is the possible mentoring role of module coordinators - and I suspect that this is the case even with fairly experienced writing teams. Writing - even of learning texts and in teams - is an intensely personal process. It is all too easy for coordinators, who may well be the most experienced writers in their teams, to focus only on the final product (and thus sometimes on their own writing) at the expense of giving full attention to the needs, developmental or otherwise, of team members.

Being reasonable people, most coordinators agree fully with the need to take this managerial function seriously. However, a wide gap may develop between principle and practice, especially if coordinators find themselves having to provide what they regard as an excessive amount of supportive and/or critical comment on the work of one or more writers in addition to completing their own writing. It thus becomes necessary to remind coordinators of this function periodically - preferably in the circulated 'newsletter' referred to above.

There is clearly a tension between a process aimed at the development of writers, and the need to deliver a quality product, cost-effectively and on time. As pointed out above, a diverse team may well contribute to the quality of the end-product in a number of ways; however, if some of the writers lack experience, the additional burden on the more experienced writers, particularly the coordinator, is likely to contribute to delay.

If collaborative writing is undertaken by a homogeneous team consisting of writers with relatively similar backgrounds, the advantages of collaborative writing are limited, and may be insufficient to offset the possible losses in time and cost efficiency often incurred in team approaches.