Conference Agenda

Networks serve as important enabling conditions for individual actors to pursue and implement changes to organizational models that diverge from existing institutions (i.e. institutional entrepreneurship). While much work on institutional entrepreneurship has considered the organizational-level of analysis, lesser attention has been paid to the role of individual actors’ embeddedness in local geographic communities as a facilitator of divergent institutional change. I examine this domain by drawing upon Obstfeld et al.’s (2020) network assembly perspective of entrepreneurial action. I leverage archival documents (e.g., correspondence, newspapers, etc.), oral histories, and secondary sources to examine a historical case of a public downtown college in the Southern United States. Largely catering to a non-traditional working adult population, institutional entrepreneurship efforts were challenged by strong political and institutional resistance to the divergent organizational form. Nonetheless, these change efforts succeeded, ultimately elevating the college’s designated status from a junior to senior institution. I argue the enterprising spirit and personal popularity of the college’s director, and later president, played a critical role in ensuring internal and community-level network resources could be mobilized, ultimately securing political and institutional legitimacy for the college. This work contributes to the literature by using a process approach to examine Obstfeld et al.’s (2020) model within the context of institutional entrepreneurship (as opposed to new ventures/start-ups). At the same time, this study considers the crucial role networks embedded within local geographic communities played in institutional change efforts, while considering how individual characteristics (e.g., personal popularity) contributed to successful network mobilization efforts.

Scientific collaboration between academia and industry is now commonplace, particularly in technical fields of research [8,6,9] such as AI [4,1]. While many sociological, economical and research policy studies highlight the motivation, profile, and impact of scientists and institutions involved in such research [8], only a few focus on the various forms of cooperation between academics and industrials.

Among the latter, we can cite the notable work by Shinn and Lamy [10], who proposed a categorization of collaboration regimes. The authors contrast the “Academic” ethos, characterized by punctual, limited, and strategic involvement, with “Pioneers” who develop long-lasting and synergetic working relationships.

However, the field of AI challenges this partition. Authors such as Ahmed et al. [1] or Jurowetzki et al. [5] show that industrials dominate the field, with scientific and technical productions centered around a few companies, and Garousi [3] highlights the difficulties of collaboration in software development.

These facts may contribute to reinforce boundaries between academia and industry.

Through an in-depth analysis of collaborative project within GitHub repositories, our study questions the relevance of the aforementioned categories in the production AI technoscientific objects and provides new insight in the way academic and industrial actors organize their collaboration.

For our initial investigations, we focus on two deep learning repositories authored by both academics and industrials: Scikit-Learn [7] and Apache-MXNet [2].

To conduct our research, we have collected extensive information on users and commit’s and issue’s history. In addition, we have retrieved institutions using the users’ profiles and a manual retrieval for the most active committers. We then have classified users as belonging to either academia, industry or a mixture of institution types. Finally, based on the commits, we have constructed time-aggregated co-modification networks for the repositories, which constitute proxies for collaboration networks.

Our analysis reveals two distinct collaboration regimes among programmers involved in the two projects.

First, the development of Scikit-Learn is almost equally shared between academicians (39%) and industrials (43%). This is shown by the inter-commit time, which follows very similar trends between the two groups, as well as the response time in issues’ discussions. Moreover, the collaboration network of the repository exhibits a significantly high probability of collaborations between users’ types. These observations testify to an equal division of technical labor in the repository.

This mixed regime contrasts with that of MXNet, which is mostly invested by industrials (70% with 27% coming from Amazon). Initially developed with an uneven distribution of collaborations since 2015, the repository experienced the withdrawal of academic contributors in 2020, resulting in full industrial management that have reinforced this asymmetry.

Overall, Scikit-Learn shows a strong academic-industrial synergy, with users embracing the “Pionners” ethos, whereas MXNet seems more fragmented in the labor distribution over the course of its development.

Although this initial research shows significant differences in collaboration regimes, further research is needed, focusing on issue discussion threads and committer diversity across files/modules. Finally, we plan to extend our investigation to other repositories to get a broader view of possible collaboration regimes.

The paper deals with the problem of interaction between conflicting orders of worth in organizations. When multiple orders of worth interface in the same setting, this state of organizational dissonance leads to situations of friction-based conflict. Friction constitutes a double-edged sword for organizations, as status competition can be. If well channelled, it can foster reflexivity and innovation enhancing corporate social capital; if mishandled, it can cause structural distress, jeopardizing corporate social capital. This research has the dual aim of exploring the relational emergence of dissonance and the collective mechanisms that mediate its impact on organizations. To this end, the paper presents an explorative case study of a non-profit organization. Using a qualitative research strategy, the interacting orders of worth are reconstructed visualizing, in a two-modes network, their links to the structural positions of their bearers while analysing the dynamics of friction within the hieratic style of the organization under investigation. Drawing on Harrison White, we conceive ties as bearing prisms of meaning as identities’ efforts at control entail their footing in specific angles of perception. Thus, the partition of organizations’ relational infrastructure in specialized substructures is linked to the emergence of differentiated orders of worth in the identities coexisting in different relational positions. Moreover, the quality of the orders of worth is reconnected to higher-level social formations addressing the organization’s embeddedness within its operational environment. Finally, the relational infrastructure is hypothesised to have a mediating role for the outcomes of dissonance through the mechanisms activated by strategic interdependence between specialised compartments.

Transitivity as a structural property of social networks reflects cohesion and can relate to the formation of social capital. In the context of migrant organizational networks, transitivity provides insight into collaborative structures, advocacy potential, and the distribution of resources. This paper explores transitivity within migrant organizational networks across five European cities -Zurich, Budapest, Barcelona, Madrid, and Athens- using binary and weighted networks to investigate cohesion and tie strength. Findings show variation among cities in terms of global cohesion, with both Athens' and Madrid's networks having higher binary transitivity, thus denoting higher levels of cohesion. In contrast, the networks in Zurich and Budapest are characterized by lower binary transitivity, suggesting more fragmented, less cohesive networks that possibly do not develop strong integrating structures. The measurement of weighted transitivity focuses on strong ties, as observed in Budapest, where fewer triads exist, nevertheless potentially fostering social capital. The results relate to the role of network structures in shaping migrant organizations’ capacity for information exchange, resource mobilization, collective and political action, and civic engagement. Thus while a high level of transitivity reflects higher levels of cohesion, the differences between binary and weighted measures suggest that it might be the strength of the ties that matters for the effectiveness of the networks. This comparative study contributes to discussions about connectivity, bonding and bridging ties, and their role in political participation. It also underlines the need for future research into how network cohesion influences the long-term sustainability and advocacy strategies of migrant voluntary organizations.

La question de la sécurité alimentaire liée au commerce des ressources halieutiques en Côte d’Ivoire demeure une problématique d’actualité. Dans le pays, la principale source de protéine d’origine animale consommée par les populations est le poisson. Dans la Sous-préfecture de Béoumi, la disponibilité de cette protéine animale est liée à la présence de cours d’eau que sont le Bandaman blanc et le kan. Le dynamisme des activités de pêche dans cette circonscription administrative suggère de s’intéresser à la contribution du commerce des ressources halieutiques à la sécurité alimentaire des populations riveraines. Cette étude vise à expliquer la contribution du commerce des ressources halieutiques à la sécurité alimentaire dans la Sous-préfecture de Béoumi. Pour y parvenir, la méthodologie déployée s’est appuyée sur l’exploitation des données issues de l’observation, la recherche documentaire, des entretiens avec les populations et les acteurs du commerce des ressources halieutiques. Il ressort de cette étude que le commerce des ressources halieutiques dans la Sous-préfecture de Béoumi contribue faiblement à la sécurité alimentaire des populations. Aussi, dans le but de réaliser un maximum de profit, les poissons sont vendus hors de ladite circonscription administrative et coûtent excessivement chers au regard des revenus faibles des populations.