The book “Innovation through Collaboration – interactive learning in nutrigenomics consortia” has been published and is also available online.

front cover


Interactive learning has been a central concept in innovation studies since the 1980s and its influence on innovation has been shown in numerous studies since then. Interactive learning is of special importance in emerging technologies in which complementary, often tacit knowledge has to be exchanged and combined in order to learn and innovate. However, sufficient insight into the black box of the interactive learning process was lacking thus far. Therefore, the purpose of our research was to improve the insight into interactive learning of heterogeneous stakeholders in the context of emerging technologies such as nutrigenomics. In line with this we formulated the following central research question:

How can interactive learning in emerging technologies be conceptualised, and how can this conceptualisation provide insights into interactive learning between heterogeneous stakeholders in nutrigenomics?

In order to answer this question we divided our research into two parts. In Part I we developed the Framework for Interactive Learning in Emerging Technologies (FILET) and in Part II we explored the FILET in two nutrigenomics consortia.

Interactive learning in emerging technologies

The concept of interactive learning in innovation processes was introduced by Lundvall in 1985 (Lundvall 1985) who later defined interactive learning as “a process in which agents communicate and even cooperate in the creation and utilisation of new economically useful knowledge” (Lundvall, Johnson et al. 2002, p226). The positive influence of interactive learning on innovation has been acknowledged in various studies (e.g. (Von Hippel 1988; Lundvall 1992; Coombs, Green et al. 2001; Smits 2002; Moors, Enzing et al. 2003; Oudshoorn and Pinch 2003; Rohracher 2005; Smits and Hertog 2007; Boon, Moors et al. 2008; Moors, Boon et al. 2008; Smits and Boon 2008; Nahuis, Moors et al. 2009)). However, detailed insight into interactive learning itself is – still – lacking. At the same time, studies on interactive learning pre-dominantly focus on the outcome of the interactive learning process, rather than what learning is and how the outcomes are achieved (Meeus and Oerlemans 2005). This black box character of the interactive learning process asks for a framework focusing not only on the interactive learning outcome but also on the interactive learning process itself. The development of such a framework and the exploration of the developed framework in two real life cases constitutes the core of this thesis.

The diffusion (Rogers 1962) or technological performance (Foster 1986) of innovations can be depicted by an S-curve with various life stages (Tidd, Bessant et al. 2001). Technologies go through several life stages from invention (the original idea) to innovation (the successful economic and/or social application of the invention in a product, process or service). “Emergence is the process or event of something coming into existence. For technological development this notion then relates to the very early stages of technological development.” (Van Merkerk and Van Lente 2005).
Knowledge is often still tacit in an emerging technology because not all new knowledge has been codified in e.g. scientific articles (Senker 1995; Arundel and Geuna 2004). Even when knowledge has been codified, it sometimes needs tacit explanation, for instance on how to perform experiments (Senker 1995; Howells 2002). Interactive learning facilitates the interchange and combination of complementary knowledge in codified and tacit form (Malmberg and Maskell 1999; Doloreux 2004). For stakeholders, interaction and collaboration is important because it is difficult for individual stakeholders to keep up with the rapid developments (Ponds 2008), especially in emerging technologies like nutrigenomics where complementary knowledge from nutrition and genomics has to be combined. The growth in scientific (sub)fields (Stichweh 1996) and interdisciplinary fields like nutrigenomics (Ponds 2008) has resulted in a division of highly specialised knowledge among heterogeneous stakeholders. The specialisation into (sub)fields has also resulted in more and more highly specialised instruments and research methods and has led to a further specialisation in know-how and expertise regarding these instruments and methods (Katz and Martin 1997). Related to the increase in specialisation are the increased costs for research. The need to combine complementary, specialised and often tacit knowledge and research methods, and the search for human resources and research funding, require collaboration of and interactive learning between stakeholders.
Interaction between stakeholders in an emerging technology is not only important for the advancement of science. In the emerging phase, the technology is still ‘fluid’ and it is difficult for stakeholders to specify desired characteristics. When the technology becomes more ‘solidified’ due to increasing vested interests, stakeholders know far better what they want, but the options to intervene decrease. This trade-off is known as the Collingridge dilemma (Collingridge 1980). Therefore, interactive learning between stakeholders is also important in the early phase of technology development for the co-construction and realisation of a shared vision among the stakeholders. Whereas knowledge is necessary to bring forth new product, process or service innovations, a shared vision can act as a driver and shared frame of reference for the further development of the technology (Checkland 1988; Vergragt 1988; Smits 2005). In a later phase of development this shared vision also will contribute to the quality and acceptance of innovations based on the technology.

To conclude, interactive learning is important in emerging technologies. Interactive learning facilitates the combination of resources and complementary (tacit) knowledge from heterogeneous stakeholders, the development of a shared vision and stimulates the advancement of science and, eventually, the development of innovations.

Framework for Interactive Learning in Emerging Technologies (FILET)

In Part I of our research we constructed the Framework for Interactive Learning in Emerging Technologies. The FILET is a heuristic that can help us to understand interactive learning in emerging technologies (Fout!Verwijzingsbron niet gevonden.). Interactive learning is a process with an outcome and the process itself is influenced by conditions.

The outcome of interactive learning can be divided into an interactive scientific knowledge outcome and the realisation of a shared vision. The realised shared vision is the result of the co-construction of a shared vision by the stakeholders at the beginning of the collaboration and the extent to which the interacting stakeholders are able to realise that shared vision.

The interactive learning outcome is influenced by the main elements of the interactive learning process. These elements include the prime mover, network formation, intermediary, and the knowledge flows between the stakeholders. The prime mover takes the initiative to bring complementary stakeholders together . During the network formation the stakeholders start to form a consortium. The knowledge flows contain the knowledge that is produced and exchanged between the stakeholders. Intermediaries can help in translating knowledge and bridging possible knowledge gaps between the individual stakeholders.

The elements of the interactive learning process in turn are influenced by a set of proximity conditions, in particular the geographical, cognitive, regulatory, cultural, social, and the organisational proximity. Geographical proximity refers to the physical distance between stakeholders and cognitive proximity to the distance between their technological foci. Regulatory proximity refers to (inter)national regulations, government funding schemes and mutual agreements that can influence the interactive learning process. Cultural proximity refers to the informal rules adhered to by the stakeholders. Social proximity indicates the presence of trust between stakeholders and organisational proximity points at the flexibility and coordination of research within the consortium.

Nutrigenomics, an emerging technology

The development of nutrigenomics has caused great expectations about future applications for personal health care. Nutrigenomics is the study of the interaction between nutrition and the genome, and envisioned applications encompass nutrigenomics health tests, dietary services and so-called functional foods with an additional health promoting benefit (Ronteltap, van Trijp et al. 2007). These applications could for example be used in the prevention and treatment of genetically predisposed nutrition-related diseases like the Metabolic Syndrome. The Metabolic Syndrome is a combination of risk factors that eventually lead to cardio-vascular disease (Kahn, Buse et al. 2005). The Metabolic Syndrome starts with obesity. Obesity is responsible for 10-13% of deaths and 2-8% of health care costs in Europe, and is, therefore, “one of the greatest public health challenges of the 21st century” (www.who.int/nutrition/topics/obesity/en/index.html 17-11-2008).

Nutrigenomics looks promising for consumers, patients and health care systems as a whole. Researchers however are still in the early phase of this innovation trajectory. At the moment they are focusing on the unravelling of the interaction between nutrition and the genome. For this research complementary knowledge from stakeholders with backgrounds in nutrition research, genetics research and the food industry has to be combined. Heterogeneous stakeholders have to interact in order to learn and innovate on the boundaries of these complementary knowledge pools. Therefore interactive learning is crucial for innovation in nutrigenomics.

Nutrigenomics is an emerging technology because it has three characteristics typical of an emerging technology:

  • There is no dominant design or definition of the technology. For our research we defined nutrigenomics as: research into the relationship between genomes, nutrition and disease (risk) and the future applications that might result from that research.
  • While there are no commercially applications derived from the emerging technology as yet, there are high expectations for future applications. Expected applications (innovations) resulting from nutrigenomics research are functional foods, nutrigenomics research as a service innovation to substantiate hard claims for functional foods based on scientific evidence, and nutrigenomics tests and dietary advice. These applications could lead to opportunities for preventing and treating nutrition-related, genetically predisposed diseases like the Metabolic Syndrome.
  • There is a perceptible increase in linkages between stakeholders in emerging technologies. For nutrigenomics, complementary stakeholders have initiated the formation of consortia in which they collaborate in nutrigenomics research.
  • Analysis of patents and publications showed that nutrigenomics is an emerging technology that attracts a great deal more attention than other technologies, as may be concluded from relative growth in number of patents (i.e. a so-called hot-spot analysis). The presence of co-publications and co-patents indicates collaboration and interactive learning within the emerging technology of nutrigenomics.

    Interactive learning in emerging technologies is a phenomenon that occurs in real life. The phenomenon is socially embedded and it is not possible to take it out of this context (as one would do in a lab experiment in a controlled environment). In these real-life phenomena the boundaries between the phenomenon and the surrounding social context often is not clear, and exogenous mechanisms might also be of importance. Case studies emphasise the rich real life of phenomena and their surrounding context. Given this real life character and the fuzzy boundaries between interactive learning and its social context, the case study becomes the preferred research strategy for the exploration of interactive learning in emerging technologies.

    We performed an exploration of the FILET in two nutrigenomics consortia selected from a list of eight consortia using specific selection criteria (i.e. similarity, interactive learning outcome, nutrigenomics as the core business, and availability and accessibility of empirical data): The Dutch Nutrigenomics Consortium (DNC) and the German Competence Network Metabolic Syndrome (CSM).

    Exploration of the FILET in Nutrigenomics consortia

    In Part II of our research we explored the FILET in the Dutch Nutrigenomics Consortium (DNC) and the German Competence Network Metabolic Syndrome (CSM). The interactive scientific knowledge outcome of the consortia included (co-) publications and standards (all codified), and an increase in tacit knowledge in the form of know-how, expertise and problem-solving capabilities.
    The shared visions that were co-constructed at the beginning of the consortia were only partly realised. Through the scientific research conducted within the consortium and the resulting interactive scientific knowledge outcome it became evident that more nutrigenomics research is needed to fully understand the intricate relationship between the genome and nutrition. Consequently the stakeholders could only realise the first objective of the shared visions (interactive, scientific knowledge) but failed to realise the second objective: useful applications.

    The main elements of the interactive learning process are the prime mover, network formation, knowledge flows and intermediay. In the case studies we observed a typical ‘sequence of events’ in the interactive learning process. First, initiated by the prime mover, a network of complementary stakeholders is formed. During the network formation a shared vision is co-constructed. Second, during research conducted within the consortia knowledge flows between the stakeholders results in interactive learning and increased scientific insight. The exchange of knowledge is sometimes facilitated by an intermediary. The scientific insights determined the extent to which the stakeholders could realise the shared vision.
    The prime movers initiated the formation of networks driven by a sense of urgency to realise promising applications and the perceived need to bring together knowledge and resources from a heterogeneous set of stakeholders. During the network formation the stakeholders co-constructed a shared vision. In the DNC case the individual visions were integrated into a shared vision with the help of an intermediary by designing interrelated work packages. Individual work packages corresponded with the expertise of the leading stakeholders, and the work packages were subsequently integrated in the ‘integrating’ work package. The formation of a network of stakeholders from nutrition and genomics fields was necessary to gain further scientific insight into nutrigenomics. In the CMS case the shared vision only was an umbrella that encompassed the stakeholders’ individual visions.

    The knowledge flows between the stakeholders resulted in an increase in interactive scientific knowledge outcome as manifested in scientific articles, and tacit know-how and expertise. During the research within the DNC, mouse models and TNO fulfilled an important role as the intermediary in this interactive learning process by functioning as a shared frame of reference, and the facilitation of exemplification projects, respectively. Ultimately however, the stakeholders were only able to realise the first objective of the shared vision. During the lifetime of the consortia it became clear that it was still too early to transform these scientific results into practical applications. To do this, more scientific questions regarding the complex relation between the genome, nutrition and disease have to be answered.

    Geographical proximity and cognitive distance had a positive influence on the prime mover, network formation and the knowledge flows. Regulatory proximity at the macro level (funding, EU regulations) positively influenced the prime mover and the network formation, and regulatory proximity at the meso level had a positive influence on the knowledge flows. Organisational and social proximity had a positive effect on the knowledge flows. For cultural proximity we have to conclude that, despite the cultural differences between research institutes and firms, the stakeholders did have a similar focus on publications and patents.

    Geographical, cognitive, cultural and regulatory proximity at the macro level were already determined before collaboration in the consortia started. Regulatory proximity at the meso level and organisational proximity were shaped during the first phases of the formation process. Trust (i.e. social proximity) was already present between stakeholders that had a shared history. However, we observed that the stakeholders with no shared history assessed the presence of trust during network formation by organising face-to-face meetings within the context of the consortium.

    The major conclusions resulting from the exploration of the FILET in the nutrigenomics case studies are:

  • Government funding was a necessary pre-condition for the nutrigenomics consortia. Without the availability of government funding the stakeholders would not have embarked on the long term, risky and expensive nutrigenomics research.
  • The case studies showed a ‘sequence of events’ in interactive learning. First, a formative stage in which the prime mover contacts complementary stakeholders and the network is formed. In the formative stage a shared vision is co-constructed by the stakeholders. Second, a research stage in which the scientific research is performed. Knowledge between the stakeholders is exchanged and combined in the research stage. This interactive learning results in an interactive scientific knowledge outcome as expressed in the scientific articles and in an increase of tacit knowledge. On the basis of these scientific insights the stakeholders can determine to what extent it is realistic to realise the shared vision they co-constructed at the beginning of the collaboration.
  • Some dimensions of proximity that influence the interactive learning process are not fully determined at the beginning of the collaboration and are further defined during the formative stage:
  • o Social proximity between stakeholders is present if they have a shared history. In the case studies not all stakeholders had a shared history and those stakeholders used the first meetings during the formative stage of the consortium to assess whether there was trust between them.
    o Regulatory proximity at the meso level is indicated by mutual agreements settling IPR and NDA issues. These mutual agreements are drafted and signed during the formative stage before the actual research collaboration between the stakeholders begins.
    o During the formative stage the stakeholders agree upon how their research will be organised within the consortium. This arrangement determines the organisational proximity.

  • In the research stage we observed that objects like mouse models also fulfilled an intermediary function. These objects acted as a common point of reference that helped the heterogeneous stakeholders to cross the boundaries between their knowledge fields (i.e. nutrition and genetics). Thus, the intermediary function can be fulfilled not only by a stakeholder, but also by ‘boundary objects’.
  • These major conclusions are incorporated in an adapted visual representation of the FILET (Figure 1).

    Figure 1 FILET based on major conclusions of case studies

    Conclusion and recommendations

    In this research a Framework for Interactive Learning in Emerging Technologies was developed. Through the exploration in case studies this framework has shown to be internally valid for the two cases studied. The concepts constitute a complete framework and the relationships were useful in describing and analysing interactive learning in the two selected nutrigenomics consortia. This research helped to take our insight into the black box of interactive learning in the development of emerging technologies one step further. Policy makers may use the results to further differentiate and focus their policies: i.e. not to imperatively require applications as an outcome of science based research, and to incorporate indicators covering tacit knowledge outcome in evaluations. For consortia stakeholders we recommend to use a checklist for the so-important consortium agreement and to build-up trust at the beginning of the formative stage. For companies we recommend to invest in R&D in order to substantiate claims for functional foods.