Page - (000136) - in Knowledge and Networks

129 approach entailed a close look at the structural factors that explain the network’s relational turnover, that is, the creation of new ties that are added to or supplant hitherto existing ones. This heuristic spinning-top model helps illustrate an OMRT context for pro- cesses such as intraorganizational learning. The dynamics of the advice network examined in this commercial court can indeed be represented intuitively as a spin- ning top. They are driven by the rotation rule in the formal structure of organization. Because judges seek advice first within their own Chamber, and because they change Chamber every year, the relational turnover in this network is high. Each year, each judge leaves behind several advisors and creates new advice ties within his or her new Chamber. This turnover, however, is compensated for by the creation of a set of advisors with epistemic status to whom judges turn for advice thanks to the Chamber in which they work. The centrality scores of members with epistemic status rise, then tend to decline over time, showing that the stabilization of this elite set of judges adds to the complexity of the dynamics of advice networks. Those dynamics come to include formally induced homophily, relational turnover, emer- gence of status as an endogenous effect reinforcing exogenously defined status, centralization of the advice network, and strategies of stabilization of this elite under capacity constraints. It is likely that empirical observation will find a perpet- ual cyclical pattern of centralization and decentralization in the advice network and that relative structural stability is achieved in part through OMRT. These detailed analyses show that most judges achieve centrality over time, some of them to the point of losing part of it and their corresponding status in the cyclical dynamics precisely because they succeeded at sharing their status by delegating a degree of their advisory function to other colleagues (Lazega et al., 2011). Lack of space in this volume precludes detailed treatment of the substantive reasons for these dynamics of the advice networks in this specific context. The complex story behind this process of collective learning is a matter of alignment with the supercen- tral judges who maintain themselves by trying to exercise epistemic control and balance excessive requests for advice (when too few colleagues occupy the top of Table 7.1 Collective learning as a cyclical process: increase, then decrease, of centralization in an advice network over time Independent variables Parameters for period 1a (Wave 1–Wave 2) Parameters for period 2b (Wave 2–Wave 3) Rate parameter 22.25 (2.03) 30.58 (3.14) Density −1.74 (0.09) −2.23 (0.18) Reciprocity 0.95 (0.16) 0.71 (0.13) Transitivity 0.50 (0.04) 0.19 (0.01) Popularity of alter 3.34 (0.40) 3.84 (0.25) Activity of alter −14.44 (1.84) −1.86 (1.87) 3-cycles-of-generalized- exchange effect −0.29 (0.09) −0.07 (0.01) Note: Adapted from Lazega et al. (2006), p.119 a N = 91. b N = 113. Standard errors are in parentheses 7 Organized Mobility and Relational Turnover as Context for Social Mechanisms…