Ten years of experience in cooperative actions for training and education in Speech Communication Sciences can be summarised as a series of achievements (see figure 1) but can also be considered from a higher perspective. We could learn about the requirements for fruitful networking such as the role of size of the network, the necessity to have limited aims and well-defined deliverables, the motivation of members to compensate for limited finances, its management. Equally interesting is the sequence of actions that can be undertaken, from practical activities, such as student exchange, to the more complicated and delicate tasks of proposals and recommendations for curricula contents.

A major requirement for a successful network is the full coverage of a community without becoming too large. This community is defined by the topic and aims of the network. The more limited the aims, the easier it is to realise a full representation of the academic community. Yet, some challenges and visions should be incorporated to make the network attractive. The cooperation in Speech Communication Sciences therefore did not start with the classical studies in Phonetic Sciences (as part of humanities faculties) but we deliberately included Spoken Language Engineering, studied in technical universities, as well. The philosophy behind this choice was that such a cooperation (both in research and education) would be the driving force behind future developments. Already implemented in 1990, this is even more valid today.

The academic community in Phonetics and Speech Communication can be estimated to constitute about 200 departments in Europe. These include departments that present a full four-year curriculum but also smaller sections that only present a few courses. The active part of the network (for Western Europe) has only about 20-30 members, but includes most countries. The full coverage of countries is very important for the understanding of all characteristics and differences between European curricula. This mutual understanding has been promoted by the realisation of a book on European Studies in Phonetics and Speech Communication (Bloothooft et al., 1995) that presents useful information for students on 169 sites, the history and present interests of Phonetics and Speech Communication per country and site, but also the views of 33 experts on major aspect of the field to make students enthusiastic, and a first description of elements that could constitute a study of Phonetics and a study of Spoken Language Engineering. Such an achievement seems to be critically dependent on a strong feeling of being a community - which relates to the limited scope of the subject - and motivated members in all countries (but a manageable number of departments per country).

Given the good cooperation shown under Erasmus, it was not difficult to create a new Socrates Thematic Network with emphasis on reflection. However, the extension with Speech and Language Therapy (SLT) in this network, although well motivated, actually brought together two different communities. This reduced the homogeneity of the network and asked for separate approaches for the different areas (this time for Phonetics and Spoken Language Engineering as well). The SLT community has no fully developed network (with electronic communication), is much bigger than Phonetics and Speech Communication, is far more heterogeneous in its curricula across Europe, has besides theoretical and experimental also therapy goals of study, and includes besides speech sciences many other disciplines. A careful choice of realistic aims in the SLT area and first steps to bring SLT closer to Phonetics and Spoken Language Engineering are required.

Clearly defined deliverables of the network, in the Thematic Network to be realised by working groups not larger than 8-12 persons, with well-kept deadlines, play a key role in its success. The cooperation in Speech Communication Sciences resulted almost every year in tangible output. During the first years there was an annual information book to support student exchange, which later was transformed into the book on European Studies. The same scheme is kept for the Socrates Thematic Network: a book every year. And it seems to work!

Long term networking allows to set new goals now and then and to touch upon increasingly difficult tasks. Moving from student exchange, we are now carefully working towards proposals and recommendations for (future) curricula. Moving from the practical exchange towards the kernel of studies is a final test of network properties. If the network addresses the very heart of the studies, the network should be fully representative and the network should have gained enough prestige. In our planning we will use a full year to raise criticism and comments on our proposals from the whole community. We always stress that our final recommendations are never obligatory and only meant as ideas that can be used to maintain and improving the studies, and as ideas that may initiate further thinking and experimenting.

The development of a European Masters in Language and Speech (technology), in another Socrates project, is such an experiment that may bring changes in our ways of cooperating. To overcome the legal barriers between countries when it comes to a Masters, the present idea is that professional organisations (the European Speech Communication Association and the European chapter of the Association for Computational Linguistics) could play an important role. On the basis of a definition of the contents of the Masters any student who acquires the necessary knowledge and skills by attending accredited courses is awarded a certificate by the professional organisations. This certificate can be awarded besides any legal degree the student may obtain from his home university. Courses can be attended across Europe and even may contain a considerable virtual component. A student should be acknowledged through the European Credit Transfer System. This would open possibilities for all European students to acquire the certificate. The prestige from the certificate should come from the professional organisation and from prospective employers who value it. No doubt that the difficult part in such a scheme is the accreditation body.

Many challenging tasks are still ahead. The continuation of networking is an essential part of it. The dangers are shown in Figure 1 which shows a clear drop in activities in the poorly financed phase between Erasmus and Socrates programmes. To assure this continuation, the network could try to raise financial support from its members and to organise itself as an independent organisation. This may be feasible for networks that unite full faculties but our field is probably too small for that. An alternative lies in an embedding in a professional organisation which takes the responsibility for continuing activities in the area of education and training. But since a network cannot survive on motivation alone and our professional organisations are poor, also in this case additional financial resources should be sought for projects.

Figure 1. A chronological overview of network activities in Speech Communication Sciences education over the past ten years. A star indicates a publication.