Astolfo

Image credit: Ludovico Ariosto, Orlando furioso (Venice: Valgrisi, 1562), p. 382. Courtesy of the George J. Mitchell Department of Special Collections & Archives, Bowdoin College Library.

Astolfo’s Network

In Ludovico Ariosto’s Orlando furioso, there has been scholarship written about themes in the poem, such as the role of gender and religions in how the characters interact with one another. As we focus on the ‘fringe’ of the poem, in this particular case being the episode that takes place on the Moon, I wanted to see how Ariosto may have tied these themes into this lunar episode through examining Ariosto’s chosen voyager: Astolfo.

Astolfo is a character that is present throughout the entire poem and interacts with many others throughout the epic. Just like his voyage to the Moon, he is also the one character that rides a hippogriff entirely around the world (unlike Ruggiero) and visits the faraway, fictional Alcina’s Island. I figured that there must be a reason that Ariosto chose Astolfo specifically for this role as a voyager into unknown, fringe areas. With all of the focus Ariosto gives to the main themes of the poem such as the interplay between different genders and religions, I wanted to see how Astolfo can be seen as an extension of these themes even to an area as far removed from the main events of the story such as the Moon.

To analyze how Astolfo is connected to these themes, my methodology involved the analysis of network data that we have formed using various indexes, including the Reynolds index data. The data was transformed into a network data set of edges and nodes for use in Gephi, a tool for network analysis. By using Gephi, I was able to isolate Astolfo and the network surrounding him, leaving me with a data set of the people and places connected to Astolfo by virtue of being mentioned in the same octave as him. Using this new data set, I performed analysis on both the edges associated with him (co-mentions) and the nodes (the characters with whom he was mentioned). To do this I entered the data laboratory in Gephi, and from there selected my desired node - Astolfo - and then selected the related edges and moved them all to a different workspace. The resulting network can be seen in the figure below. From there I was able to find the proportions of things such as the gender and inferred ethnicity of those characters that are closest to Astolfo. The variables being examined here are features that we gave to the characters through the use of various indices, including implied gender and origin for the nodes in the network.

Astolfo's Network

Astolfo is the large, purple node near the center of this graph.

Astolfo's Network Statistics - Gender

First, comparing the overall number of men and women in the overall index to the people found solely in Astolfo’s network, there is a three percent difference. That difference is only marginal, but by examining the target of edges, which lets us see who is actually being interacted with, I have found that a higher amount of women are being interacted with in Astolfo’s network than the overall index. What this tells us is that while there are not very many more women in the network as compared to the character list as a whole, the women in Astolfo’s network are disproportionately the targets of edges, showing that they may be more prevalent in the characters that Astolfo interacts with than the rest of the characters in the poem.

Astolfo's Network Statistics - Origins

Similarly, by examining the religions of sources of nodes in Astolfo’s network, we can see that when comparing the data to the overall index, there are roughly three percent fewer Western Europeans, with most of the other data categories being roughly equivalent. But through examining the sources and targets of the edges in the network, it can be seen that Western Europeans make up ten percent fewer of the targets than of the edges, showing that while many Western Europeans are interacting, it is not always with each other. As these are all of the people in Astolfo’s network, it can be seen that Arisoto may have been using the people associated with him to broaden the scope of the poem, through having these Western Europeans associated with him interact with many people from different religions.

Orlando's Network

I constructed a similar network associated with Orlando, which can be seen to the left, in order to compare the results from Astolfo’s network with another character in the poem. I chose Orlando as he is the namesake of the poem and I thought he could be representative of the poem as a whole. He is represented by the purple node in the center of the graph on the left.

Orlando's Network Statistics - Gender

In comparison to Astolfo’s network, his network features similar amounts of males but a lower number of women and a higher number of characters defined as ‘other.’ In the sourceimpliedgender, his network is similar to Astolfo’s in terms of women, but has fewer men and more other characters. Otherwise, the most stark difference is the percentage of women in targetimpliedgender, which has only 11.36% women in comparison to his 82.19% men.

Orlando's Network Statistics - Origins

In terms of different ethnicities found in these networks, Orlando’s network is 28.32% Western European, with Africans being 12.32%, and Asian characters being 7.96% of the nodes found in the network. Astolfo’s network, on the other hand, features only 21.4% Western European, 11.4% African, and 8.5% Asian characters. This is a large disparity, as well as the fact Astolfo’s network features far more characters described as ‘African/Greek’ (primarily Ruggiero and Marfisa). In Orlando’s network, 33.37% of edges have Western European sources, which is roughly seven percent less than Astolfo. Orlando’s network also features 35.21% of edge targets being Western European, a whole 5 percent more than Astolfo.

A Statistical Analysis

The statistics reported above are not incredibly different. I used a chi-squared test to determine whether the statistics for Astolfo and Orlando are independent from the total statistics of the entire network of the poem. To do this I first found the total percentages of the variables being measured for the total index, Astolfo’s network, and Orlando’s network. Then I used the =chisq.test function in Google Sheets to calculate the p-value that measures correlation between the two, which can be seen in the figures below.

Combined Stats - Gender

Using the percentages from the total index as the expected values in this calculation, I then found the relation between the incidence of gender in the nodes found in each network. I found that in terms of gender, the nodes found in Astolfo’s network are not significantly related to the overall index of nodes in the index. When comparing the percentages for male, female, and other in Astolfo’s network to the overall network, the p-value returned is 0.1969697423. Using a standard alpha of 0.05 shows that this value is insignificant. Then I repeated the same process for the percentages of gender in sourceimpliedgender and targetimplied gender. The values returned for Astolfo are 0.9262124453 and 0.2948725106 respectively, showing that they are insignificantly related to the overall network. Whereas for Orlando, when the same process was repeated for the nodes in his network, the value returned was 0.0007948111552. This shows a high significance in the relation to the overall network. When computing for sourceimpliedgender and targetimpliedgender in the edges, the values 0.00001986916402 and 0.01375387117 are returned - both being well below 0.05 and showing high significance.

Combined Stats - Origins

I repeated this same process with the implied religions of characters to learn how Astolfo’s network differs from both Orlando’s and the overall network, which can be seen in the figure below. The results here were not as striking as when comparing by gender. When entering the variables found for our personorigin variables such as WesternEuropean, African, Greek, Asian, etc. there is little difference found between Astolfo’s network and Orlando’s or the overall network. When comparing Astolfo’s network to the overall network, I found that the comparison between the overall network for personorigin gave a chi-squared statistic of 0, showing there is no observed difference. When looking at the edges, in sourceimpliedreligion and targetimpliedreligion the values 0.007469575772 and 0.005269447009 were returned. When repeating the same process for Orlando, the comparison of the personorigin returned a chi-squared statistic of 0.000000006033440192 and the sourceimpliedreligion and targetimpliedreligion returned 0.000001302282082 and 0 respectively.

These results show that Astolfo’s network is greatly different from the overall network of all people in the poem. When comparing gender in Astolfo’s network to the overall network of characters in the poem with a chi-squared test, high chi-squared statistics are found. This shows that there is little correlation between gender in the overall network and in the network of characters Astolfo interacts with. What can be gleaned from this is that gender operates differently in Astolfo’s network compared to the rest of the poem. Women are a greater percentage of the targets of edges, showing that women are actually being interacted with as characters, and not only acting as characters that exist to interact with men. I used Orlando as a comparison, and in his results there were no significantly different chi-squared statistics found. In my analysis of the networks in the scope of character origin and religion, I repeated the same aforementioned process done with the gender, sourceimpliedgender, and targetimpliedgender with the personorigin, sourceimpliedreligion, and targetimplied religion variables. From this analysis I did not find any significant difference in the percentages of religions in Astolfo’s network compared to the overall network of the Orlando Furioso.

In light of my question regarding how major themes of the poem, namely interplay between genders and religious backgrounds may fit into Astolfo’s network, I found that Astolfo’s network is dissimilar from the overall network of the poem in terms of the genders of the people found within the network, but overall very similar in terms of the religion and origin of people in the poem. But it should be noted that not all interactions between characters in a poem can be expressed numerically, and it would be an interesting study to delve deeper into the actual interactions Astolfo has with women or people of different faiths with tools such as sentiment analysis.

Next: Minor Characters on the Moon

Bibliography

Page updated

Report abuse