Networks of excellence promote excellence and integration. Integration is a fuzzy notion that, for us, can only be understood in a framework where competition is the prime mover for new knowledge to be produced. Integration is thus first and foremost deploying a space which adds value to our basic production structures. PRIME does so by enabling to develop "facilities" which we cannot develop individually, and by helping to foster complementarities required by new approaches or associated with problem solving. We do this from a pragmatic standpoint, that is when observing de facto "gaps" or "failures".
This was for instance the case for training and for indicators (see respective sections). But this was also the case for two other functions, for anticipating on the long-term research agenda (most teams are driven by the short or at best medium term contracts from policymakers), and for establishing a space where heterodox ideas, approaches and methods can be tested, enabling them to enter the existing funding alleys (which, in our field, remain anchored in present-day problems and dominant paradigms). We have thus developed an ‘experiment’ that helps us further investigate the ‘why, what and how’ of it.
The platform we have developed comprises four dimensions. (a) a governance structure that encapsulates a vision, organises as in any structure, 4 or 5 year plans, and provides a working operational management; (b) a space where periodic exchanges can take place and which is a cornerstone to the ability of the community to define and update its research agenda, and which is flexible enough to accommodate for targeted reflection; (c) a set of instruments that give flesh to joint activities, and (d) a reflexive capacity through the creation of an autonomous ‘characterisation group’. The following paragraphs present briefly these four dimensions of integration.

(a) Governance structures (for more detail click here). We have considered that the 2-year preparation work for the NoE provided both the vision and the 5-year ‘objectives’ (which were inscribed in the EC contract). We have organised a ‘Governing Board’ made of all partners with votes depending on members' excellence in the field and involvement in the activities. The complex voting procedure (inscribed in the Consortium Agreement and guaranteed by the Characterisation Group) has worked well, and for instance has enabled to renew by 40% the Executive Committee at mid-term. We have delegated all operational power to the elected Executive Committee (ExC) made of 12 members. The ExC meets at least 3 times a year and also works on an e-mode. Its mode of operation is to arrive to a joint agreement (with very few cases where only a majority vote was required). One central function is to allocate funds and to monitor developments. The choice made has been to allocate all funds on the base of projects bounded in composition and time, so as to insure that all funds go to activities (meaning that ‘dormant’ members get not support at all) and that activities can be redirected or stopped if considered necessary (at the time where they ask for fund renewal). Furthermore, all funds granted to scientific projects are examined by members of the Scientific Committee before the ExC makes any decision. These developments have proven efficient and lean in the consumption of the time of ExC members (who operate on a voluntary basis with only marginal support to their involvement).

(b) The annual conferences are the central meeting place of Prime participants with between 150 and 200 participants every year. In order to accommodate with the evolution of the NoE, their organisation and content has changed every year. The conference is also a place where we discuss strategy and new explorations required. The conference was for instance instrumental in the small experimentation we did about development paths in Africa (which has had recent echoes in UNESCO activities, both on activities on innovation systems see the 2007 conference on the topic, and at the high level expert meeting organised in Manchester – Feb 2008). It was also central to the reshaping of WP2 and the launch of the ERA dynamics project, and played an important role in pushing for the development of new approaches for indicators. We however recognised early on that this was not enough for discussing in depth specific evolutions of the research agenda. This is why we have developed what we call ‘review and initiation’ activities to foster states of the art and worldwide expert gathering on specific issues (Innomil on defence or Globpol on the internationalisation of firm RDI activities are exemplary of this approach).

(c) The platform cannot be lasting if it does not address the needs of the community. We have identified 3 such needs and have developed corresponding activities that have been implemented through the WP mentioned above. Two are ‘structural’ and deal with training (in a small community at the encounter of 4 disciplines, see training section) and with indicators (see our progressive construction of the ENID network). The third one was oriented toward the need for increased methodological and theoretical variety (an issue raised in our assessment of the situation in 2003). It entails risk taking on new heterodox concepts or methods. Taking risks means that some are more successful than others, but we consider that all are worthwhile since they enable learning to take place. All activities undertaken are presented in the project section. The progressive management by the ExC, by steps of no more than 18 months, has enabled to reorient or stop the less productive projects and to focus resources on projects which demonstrated their relevance both in attracting new participants and in delivering outputs recognised by the wider community (thus our emphasis on articles in journals). This is a powerful way to nurture long-term co-operations and alliances, an important aspect of our vision about the evolution of the community.

(d) A complementary feature of our network is that it deals with policies. Studying new instruments, their deployment and their effects is thus part of our remit. This is why we decided from the start to consider Prime as an experiment to be observed and accompanied. For doing so we created an autonomous ‘Characterisation Group’ of 3 persons which was in charge to develop a conceptual framework for the analysis of NoEs (see paper published in SPP, Luukkonen, Nedeva and Barré, 2006) and to monitor on-going developments. One aspect was to follow cognitive and social changes between two points (end of first year, end of fourth year). The final report of the Caracterisation Group  highlights them. Such a space has both direct and indirect effects. Direct in the activities it promotes (for instance the mobility of PhD students or the research explorations made) and indirect through the circulation of staff (for instance joint positions as this of Larédo between Paris and Manchester; ‘international’ recruitments as we can see in the case of Manchester or Twente drawing upon German and US labs; or CSIC drawing upon SPRU, or Swedish colleagues recruiting young researchers trained in Spain…) or through the visibility it gives to the field and the structuring it fosters. This is very visible for instance in France (with the creation of IFRIS¹), in Spain with the internal reorganisation of CSIC or in Portugal with the creation of the NONIUS network (see its first conference, November 2007). Furthermore two institutions are discussing ways in which to foster the emergence of a ‘knowledge community’ (of course open to others) as a way to prepare post-Prime times (CNRS and CSIC).
This tells that the long-term vision put forward at the beginning of PRIME (and aimed at materialising within a decade, not 5 years), is starting to take shape.