David Meurer

Figure 1. Navigational architecture at 2020, after 20 of 121 testimonials. Testimonials are labelled by narrator (rather than testimonial number). Seven characters provide one testimonial each. Five characters provide two testimonials each. One character provides three testimonials. The graph was generated using Force Atlas layout. Node size and colour are degree weighted (i.e., based on quantity of inbound and outbound links edges).

 

Figure 2. Navigational architecture at 2033, after 60 of 121 testimonials. Testimonial nodes are labelled by narrator (rather than testimonial number). Node density is indicated by size and colour of the node. Clusters are forming due to the higher incidence of links between testimonials provided by the same narrator. The most prominent form clusters for narrators with ve or more testimonials. The graph was generated using Force Atlas layout. Node size and colour are degree weighted (i.e., based on quantity of inbound and outbound links edges).

 

Figure 3. Navigational architecture at 2041 (conclusion), after all testimonials are complete. Testimonial nodes are labelled by narrator (rather than testimonial number). Node density is indicated by size and colour of the node. Clusters are now well formed for narrators with the greatest number of testimonials. The graph was generated using Force Atlas layout. Node size and colour are degree weighted (i.e., based on quantity of inbound and outbound links edges).

 

Figure 4. Character network at 2020, after 20 of 121 testimonials. Characters are organized into distinct and separate communities. Force Atlas 2 layout partitioned by modularity class (community detection). Node size is degree weighted (i.e., based on quantity of inbound and outbound links edges).

 

Figure 5. Character network by type at 2020, after 20 of 121 testimonials. Silent characters appear red. Force Atlas 2 layout partitioned by modularity class (community detection). Node size is degree weighted (i.e., based on quantity of inbound and outbound links edges). Node colour indicates type (silent or non-silent).

 

Figure 6. Character network at 2033, after 60 of 121 testimonials. Characters are organized into distinct and separate communities, but there are now a significant quantity of edges connecting most communities together within the network. Force Atlas 2 layout partitioned by modularity class (community detection). Node size is degree weighted (i.e., based on quantity of inbound and outbound links edges), colour indicates modularity class. Edges are weighted, colour is a mix of source and target node colour. Therefore, edges consistent in colour with node are internal to the community. Edges with mixed colour link distinct communities together.

 

Figure 7. Character network at 2041, after completion of all 121 testimonials. Communities connecting characters are now much more densely connected. Force Atlas 2 layout partitioned by modularity class (community detection). Node size is degree weighted (i.e., based on quantity of inbound and outbound links edges).

 

Figure 8. Character network at 2041, after completion of all 121 testimonials. Note the quantity of minor characters positioned outside the main clusters of principal characters. Fruchterman Rheingold layout dispersed to ll a circle. Node size is degree weighted (i.e., based on quantity of inbound and outbound links edges).