Perception and the search for humanity on The Island of Doctor Moreau

A literary network visualization by Daniel M. Wescovich

Introduction

The Island of Doctor Moreau is a weird little book. It tells the story of Edward Prendick, an English gentleman who is shipwrecked off the coast of Peru and rescued by a ship delivering supplies to an island where unusual experiments are taking place. Dr. Moreau is a vivisectionist, the practice of performing experimental surgery on live animals. Specifically, the doctor grafts human genetic material into live animals, thus giving them the unmistakable, but rather disturbing, mark of the human being. Prendick's discovery of the true nature of these projects eventually leads to the social collapse and physical death of all of the creatures of the island.

My project aims to construct several network visualizations that will map the social, physical, and perceptual relationships between the human beings and the hybrid beings (the Beast-folk) who coexist on the island. While initial collection of physical data suggesting genetic relationships among the islands inhabitants fell rather flat, the process of collecting and manipulating the data led to a far more interesting discovery.

The culminating network visualization takes every instance of the word "beast" as Prendick, Moreau, and Montgomery (Moreau's assistant) use the term to describe the hybrid creatures who live on the island. In the end, the data suggests that while all of the human beings describe the hybrids as beasts, Prendick eventually allows the term to apply to both hybrids and humans, a discovery that suggests a key feature of Prendick's disgust of Moreau's morbid creations.

I have taken measures to create a sort of meta-narrative as a vehicle for my research and findings. Key among my concerns is an attempt to place the raw data of my study in the context of narrative itself. It was only near the end of this project that I realized that this kind of meta-narrative (a data visualization project about making a data visualization project, if you will) is something I used to do for my students in order to help them think about how to complete a complex project. So, as I reflect on the winding narrative of my own work, I can't help but notice that it might actually be a project more about learning how to use new data methods to interpret literary texts, than it is a full-scale literary interpretation. Either way, I find that my knowledge of digital humanities made great strides through the completion of this project. That said, let's begin.

Beginnings

This project begins with a comment from two of my classmates on their own project. When discussing their initial impressions of the novel they chose for the project (The Invisible Man, no less, another famous piece by H. G. Wells), my colleagues noticed that the narrator never identifies how Griffin attains funding for his research into the optics of invisibility. They suggested that tracking money through the novel might be an interesting path forward.

Earlier, in my reading and research, I had taken note of an article on postcolonial themes in Doctor Moreau which highlighted how the animals the weird doctor used in his vivisections were imported from other places throughout the British Empire. The central point was that before the animals could become "humanized" through scientific experimentation, they had to be transformed into "trade goods" (Taneja 140).

So, the money-tracking idea had me thinking about sources, and the Taneja article brought to my attention the sources of the animals for Moreau's experiments. Then, I went back to the text and noticed that Dr. Moreau's rather detailed explanation of his experiments never identifies where he gets the human material that he grafts into the animals in order to humanize them. Weird, right?

Initial questions

Where does Dr. Moreau get the human material for his experiments? My initial readings would suggest several possibilities. If Dr. Moreau spliced his own human material, then that might help to explain the authoritarian paternal relationship between the doctor and his hybrid beings. However, he could have taken material from Montgomery, his assistant. This option is supported by Montgomery's conflicted affection for the hybrids.

But, what kind of data could I mine from the text that would suggest the source of human material? In a way, I was starting to think of this project as a kind of literary paternity test. Might it be possible to identify relationships among characters by analyzing physical and behavioral traits shared among them? Could a network visualization connecting the various hybrid and human beings described in the text through their shared traits indicate genetic relationships, thus revealing the source of the human material Dr. Moreau spliced into his animal subjects?

Perhaps?

The first data visualization

I knew that for this project I wanted to do a network visualization with Gephi because I found that my previous work with this tool on Bram Stoker's Dracula yielded new ways of interpreting relationships that I had not expected. I also wanted to expand my knowledge of how to use the various capabilities of the software.

So, I began to design a spreadsheet that would list each character and the specifically described traits ascribed to them as nodes, and another spreadsheet listing each instance of a trait associated with a specific character as edges. The hypothesis was that if I collected data from passages in the text that provided rich description of the characters in terms of their physicality and behavior, then a visualization of that data would suggest relationships among them. A hybrid and a human who shared many traits would identify a possible genetic relationship.

Wait... this isn't that interesting, or is it?

I didn't get too far before I began to see that there were some things I didn't like about this first visualization. Even though I only entered in a few select sets of descriptive data, the relationships and indications I was expecting weren't materializing. I thought that by entering data on the specific physical and behavioral traits, along with the more prominent hybrid and human characters with which those traits were associated, would result in a visualization that could indicate some kind of genetic relationship between the hybrids and one, or more, humans. The original question was where Dr. Moreau got the human material for his vivisections. It was clear that he was importing the animals from any number of different locations spread across the British Empire, but their physical and behavioral traits were not aligning with any particular human. So, what we can see in the visualization above are sets of traits clustered around each of the characters, but not much indicating any kind of relationship.

Another thing that was hovering around in my mind was an almost off-handed observation from the professor that my original way of setting up nodes and edges in the Gephi spreadsheets was not specifically charting any kind of real relationship. The more I thought about this, and the more I looked at the first data visualization, I started to understand the implications of that observation. I was charting data specific to individual characters before I had any idea of what relationships were already present. I was starting with data looking for a relationship, instead of starting with a relationship then charting data to discover deeper implications about those relationships. An adjustment of methodology was in order.

Towards a real network visualization

One of the unexpected findings from the first visualization, regardless of its flaws, was that the one concept that did in fact connect all five of the selected characters was the idea of being beastly in some way. In the data set for the first visualization each of the five characters is described as being "like a beast" (Wells 40), or having "the unmistakable mark of the beast" (42), or something else related to a beast-like quality.

When we take into account that Doctor Moreau is told largely from the perspective of Edward Prendick, the "private gentleman" (5), English, rescued from a shipwreck off the coast of Peru and taken to the strange island-home-laboratory of an outcast vivisectionist, the thread of a relationship emerges. Prendick is the one most often assigning the status of beast upon both the humans and the hybrid beings he encounters in his adventure. Could this mean that it's possible to map a relational network through the eyes of Prendick, based upon whom he most views as a beast? Could it be possible to model the same network from the view of Dr. Moreau?

A new question for new data

The question, then, becomes one of understanding how Prendick relates to others. Does he relate to them as a human to an animal, or as a human to another human being? Who else exhibits a tendency to define others according to their sense of the beastly? Since the first data visualization indicated that the view of someone being a beast was distributed across both humans and hybrids, the question of what actually distinguishes hybrids from humans, if it is not beastliness, must also come to bear on the next visualization.

It's interesting how the original question of where Dr. Moreau gets the physical human material for his experiments does not emerge from the textual data, but upon scrutiny reveals a deeper philosophical question regarding the meaning of being a human person, the relationship between humans and the animal world, and the specific perception of each.

The context of this question is within the late-nineteenth century conversation about the ethics of vivisection, the practice of performing surgery or other invasive medical procedures upon live animals toward the end of scientific advancement. Wells himself argues that anti-vivisectionists do not "clearly separate [animals and their experience] from humanity" ("Popular Feeling" 269). In light of this statement, emerging patterns of the use of the word "beast" to describe others, both human and hybrid, might allow for an interpretation into the data in terms of an anti- or pro-vivisectionist stance. Initially, I expect to find that Edward Prendick will use the word indiscriminately since he very much deplores Dr. Moreau's work, while I expect that Dr. Moreau will reserve the use of the word only to hybrids. Let's see what the data brings.

Use of the word "beast"

This second visualization tracked the use of the word "beast" through the entire text of Doctor Moreau, and several observations stood out. First, only humans use the word; hybrids do not use it at all. This suggests that while the human characters are rather keen to define the hybrids as something Other, the hybrids actually see themselves as human so it makes no sense for them to make a distinction. Second, Edward Prendick is the only human character to use the term to describe both humans, hybrids, and animals. Wells's assertion that anti-vivisectionists did not seperate the experiences of animals and of humanity seems to come through in reverse, for Prendick cannot seem to distinguish anything apart from the beastly experience. Paradoxically, Prendick, being disgusted by Moreau's experiments, comes to the same conclusion: sharing in either in the beastly or human natures results in a sense of kinship with the other that results in a need to extricate the other from that shared nature.

In other words, Prendick's perception that humans and hybrids both share an animal nature makes him much more hostile toward them in the end. This perspective leads to an anti-vivisectionist tendency, just as the perception of a shared human nature would do the same.

In conclusion

In the end, I think the larger question I have been after through this entire process is about what kinds of new readings are possible in data mining literary texts. Approaching this project from that particular space significantly impacted the way I read this novel in several ways. On one hand, I began my reading with an eye to the kinds of data I could collect, so I was rather attuned to seeing certain quantifiable elements. The scientific (as in science-fiction) themes of genetic splicing and experimentation made it easy to gravitate initially toward genetic data. However, the results led me to a very different place. The relationships indicated by the way people thought of others in terms of being a human or an animal produced data that leads right to the very heart of where we think the human ends and something else begins.

Google Sites

Report abuse