CWRL logo
EJournal Home Page
Vol. 3: Contents

Computers, Writing, Rhetoric and Literature

Andrew R. Cline
Graduate Student in Rhetoric & Composition
University of Missouri - Kansas City

Counting on E-texts for Rhetorical Analysis: The Case of Hawthorne's Influence on the Composition of Moby-Dick

Link to our discussion forum on Issues in Electronic and Print-Based Literary Scholarship
  1. The look Dr. Craig Bartholomaus gave me is an amusing memory. I had just finished showing him several pages of scribbled notes heavily marked with exaggerated arrows, connecting lines, and multi-colored highlights. These pages looked more like a Jackson Pollack painting than literary research. I explained, somewhat excitedly, that I could prove that Nathaniel Hawthorne had a profound effect on the composition of Moby-Dick at the sentence level. It was in the numbers. Melville's average words-per-sentence counts skyrocketed after chapter 22, after he met, and began to read, Hawthorne. The key was in his increased use of subordination and the proof was in the numbers. Bartholomaus blinked, smiled, and said, "I don't enjoy this kind of work, but I'm sure glad there are people who do."

  2. What he was responding to took a moment to sink in. "No, Craig. I'm not counting by hand," I said. "I'm doing it on my computer." I explained that I had obtained a copy of Moby-Dick, an electronic text, from the Gutenberg Project on the Internet. With this ASCII version of Melville's text, I was able to break the text into individual chapter files and then, using a standard ASCII text editor, generate a report on each chapter that gave me character, word, sentence, and paragraph counts. The entire operation, from downloading the text to running the final reports, took about four hours.

  3. In a way, it had been Craig's idea. Based on a review I wrote of A Thick and Darksome Veil by Thomas R. Moore--a rhetorical analysis of Nathaniel Hawthorne's sketches and prefaces--Craig suggested to me that perhaps Hawthorne also influenced the composition of Moby-Dick at the sentence level, but he was unsure how to prove it. Both authors seemed to hide certain culturally-variant themes in their subtexts--in the subordination of their sentences. It has long been accepted that Hawthorne influenced the younger Melville at the thematic level. Could I expand on this?

  4. To prove that Melville was using the same rhetorical techniques as Hawthorne would require somehow proving that Melville changed the way he used subordination after chapter 22 of Moby-Dick. It would require showing he used subordination in ways, and for purposes, similar to Hawthorne. And it would require showing that he used subordination in the first 22 chapters in a way similar to his own earlier works.

  5. My thinking went like this: Heavily-subordinated sentences are long. If Melville increased his use of heavily-subordinated sentences after chapter 22, then I should count the words and sentences in all the chapters to arrive at an average words-per-sentence count. An increase in this count after chapter 22 would be the first clue. Then, as Moore did for Hawthorne, I would put Melville's sentences to the same rhetorical analysis and show that Melville's culturally-variant themes occur in chapters with the highest counts and in heavily-subordinated sentences of a type not used in the first 22 chapters or in earlier works.

  6. If I had the patience of a Medieval monk, I would still be counting words and sentences. Thanks to computers and the scholars at the Gutenberg Project, however, I completed this work in four hours and can now contend, with a great deal of confidence, that Hawthorne had a profound effect on the composition of Moby-Dick at the sentence level. So, the fact that an electronic text of Moby-Dick exists (and Typee which I analyzed as the earlier work) and that I could easily find and obtain it made this work possible.

  7. From this research, I wrote The Scared White Doe: Glimpses of Hawthorne's Influence on the Composition of Moby-Dick. This text includes a table of data I generated using the Gutenberg Project text.

  8. John Slatin, in paragraph 45 of his chapter "La Zambinella Meets the Cyborg," discusses the nature of ASCII texts and the MLA's concern about their use in literary research and pedagogy. He writes:

    Many of the e-texts available on the Internet, including those produced under the informal aegis of Project Gutenberg, are no more than ASCII files that do not carry with them information about either their origins or their structure; impossible to catalogue with any precision, they are inadequate for most scholarly and many pedagogical purposes, as the MLA statement discussed earlier makes abundantly clear. For scholarly (and other) purposes, theories and conventions as to what constitute the important features of text are codified and embodied in information structures along with the text itself. In encoding schemes based on SGML, such as the Text Encoding Initiative Guidelines or Hypertext Markup Language for the World Wide Web, these structures are called Document Type Definitions, or DTDs. The DTD in turn regulates the behavior of other information structures, called programs, which manipulate textual elements composed of the same "material."

  9. I disagree with the MLA criticism, as characterized by Slatin, that ASCII texts "are inadequate for most scholarly...purposes." This seems too broad an indictment of this new way to manifest a text. Indeed, had the Gutenberg Project text of Moby-Dick come burdened with "information structures" other than the pure text (such as digitized images of original pages as described in paragraph 33 of Slatin's text), my analysis could have taken much longer to complete or been nearly impossible. In other words, I could still be hunched, monk-like, over my computer screen trying to cleanse Moby-Dick of superfluous information.

  10. Moby-Dick is the text itself. And ASCII is text in its purest form and lends itself to rather easy analysis by computer. Computers reading ASCII text open up a new world of textual analysis, especially for certain rhetoricians, such as myself, who are not concerned about the "theories and conventions as to what constitute the important features of text" being "codified and embodied in information structures along with the text itself."

  11. I have no reverence for ASCII text as a manifestation of literature. Instead, I appreciate it for the work it allows me to do and the information it allows me to uncover. And when I am finished with it, I simply and unceremoniously drag the file to the trash and forget about it.