Fig. 4.1. Top five words with the strongest markers of an upward trend.

Fig. 4.2. Top five words with the strongest markers of a downward trend.

Fig. 4.3. “Works” compared to “performance” and “show.”

Fig. 4.4. Changes in the relative frequency of all mentions of “audience,” and trends in purely rhetoric usages of “audience.”

Fig. 5.1. The number of connected components over time.

Fig. 5.2. The number of connected components over time for cast members.

Fig. 5.3. The number of connected components over time for production crew members.

Fig. 5.4. The number of nodes over time separated by cast and production crew.

Fig. 5.5. The number of edges over time separated by cast and production crew.

Fig. 5.6. A pairwise matrix comparing the network-theoretical measurements of Javanese and Indian characters.

Fig. 6.1. Nonidentical pixels in subsequent images as a function of time. Adapted from Escobar Varela and Parikesit (2017).

Fig. 6.2. The standard deviation and mean of nonidentical pixels for each scene, grouped by scene type.

Fig. 6.3. Combined violinplots and bar charts. The violinplots show the distribution of nonidentical pixels within each scene (the mean is indicated by a white circle within each plot). The bars show the length of each scene.

Fig. 6.4. A kayon puppet, from the front side.

Fig. 6.5. A kayon plot. Similar to figure 6.3 but the y-axis of the lower half is inverted, and the length of each scene is represented by a line rather than by a bar.

Fig. 6.6. The theoretical location of all subtypes except Jatayu along the strong vs. refined, and proud vs. humble axes.

Fig. 7.1. Scatter plots with regression lines for average number of performances per month compared to population density, land area, and total population per regency.

Fig. 7.2. A choropleth map of Java, with districts colored by average number of performances per month (grouped into deciles).

Fig. 7.3. A scatterplot of each regency’s total number of performances (x axis) against the average value of its neighbors, also known as its spatial lag (y axis).

Fig. 7.4. The results of 999 random permutations of the data. The shaded area shows the kernel density estimate of these results, and the vertical dashed line shows the actual observed Moran’s I of 0.173.