Bright Spots and Fading Lights: Dynamics of Knowledge Spillovers and Geographic (Dis)Association in the Florence Fashion Shows, 1951–84
研究意大利佛罗伦萨时装展（1951-84）如何通过知识溢出与符号溢出，塑造并最终侵蚀了地方与产业的关联，揭示了临时集群中精英参与者对符号价值的决定性作用。
This article investigates how temporary clusters forge, stabilize, and sometimes erode symbolic associations between places and industries, focusing on the lifecycle of Italy’s first international fashion shows, the Pitti shows in Florence (1951–84). Bridging research on geographic (dis)associations and temporary and cyclical clusters, we develop a framework linking knowledge spillovers to image spillovers. We argue that knowledge generated within temporary clusters translates into durable geographic associations only when selectively mediated, narratively anchored, and sustained by high-status participants. Drawing on longitudinal archival research combining participation data and systematic trade press coverage, we reconstruct how the Florence shows initially consolidated Italian fashion as an exportable category, yet later experienced symbolic hollowing. We demonstrate, first, that spillover capacity within temporary clusters is structurally asymmetric: A small number of elite actors disproportionately shape both knowledge release and symbolic value. Second, we show that these actors perform a metonymic function, standing in for broader geographic imaginaries. Third, we theorize temporary clusters as status arenas embedded in hierarchical cyclical event circuits, where centrality is dynamically renegotiated and participant mobility unevenly distributed. Fourth, we highlight the role of symbolic governance and curatorial selectivity, showing how growth without coherence can dilute credibility and accelerate reputational decline. Conceptually, the study advances cultural economic geography by theorizing temporary clusters as historically contingent sites of knowledge and symbolic value production, central to geographic associations yet reliant on fragile alignment among organizers, cultural intermediaries, and high-status participants. Methodologically, we show how historical-geographic reconstruction reveals slow symbolic dynamics obscured in cross-sectional analyses.