Page - (000159) - in Knowledge and Networks

152 ness centrality. It is worth mentioning them by name: Benny Goodman, Henry “Red” Allen, Sidney Bechet, Don Byas, and Woody Herman. These are founders of swing music who were able to survive in the network. In contrast to Louis Armstrong, who knew how to reposition himself at the core, these players remained on the outskirts of the core. Although not always in the core spotlight, recording with these players was just as coveted as playing with Elite musicians. Cluster 6: Up-Starters. This is the second largest group (725 musicians), with an average lifespan similar to Shooting Stars. However, their average betweenness and membership in the core are low (0.13 and 0.21, respectively), higher only than the Wanna-be cluster. I identified important musicians in this cluster, includ- ing the likes of Dave Brubeck and Quincy Jones. Nonetheless, this may indicate that musicians who were commercially successful were not necessarily active at the core of the jazz community during their heyday. I would like to point out some potential issues with the Elite and Shooting Stars clusters. Both groups spent an average 60 % of their lifespans in the core, a statistic that suggests the existence of different generations. Failure to distinguish between generations might lead to poorer comprehension of the trajectory of different musi- cians. For this reason, I split these two clusters into two generations using a 2-mode faction search. The Elite first generation spans from 1930 to 1944 (Cluster 31), while the second generation spans from 1945 to 1969 (Cluster 32). The Shooting Star first generation runs from 1930 to 1949 (Cluster 41), while the second generation runs from 1950 to 1969 (Cluster 42). See Table 8.3 for descriptive statistics on these new clusters. Table 8.4 summarizes the evolution of musicians by each trajectory type. Blockmodeling the Jazz Field In order to obtain a topology of the jazz field structure over time, I applied a block- modeling approach from the social networks analysis tradition. The use of block- modeling has become widespread in organizational research. DiMaggio (1986) was Table 8.3 Clusters Obtained by Splitting Elite and Shooting Star Groups Clusters Key trajectory variables 31 32 41 42 Average betweenness centrality 1.29 0.84 0.51 0.52 Lifespan (periods) 7.14 6.41 3.74 3.94 Periods at the core 4.00 3.84 2.15 2.30 Number of individuals 22 49 39 89 Note. Cluster 3 (Elite) split into first generation (1930–1944) Cluster 31 (Elite-1) and second gen- eration (1945–1969) Cluster 32 (Elite-2). Cluster 4 (Shooting Star) split into first generation (1930–1949) Cluster 41 (Shooting Star-1) and second generation (1950–1969) Cluster 42 (Shooting Star-2). (Design by the author) C. Kirschbaum