Skip to main content

Exercise 8

2 min read

I have been thinking a lot about novels’ consciousness and narration as I have been planning my final paper, and in particular I have been concerned with shifts in these areas in novels at different times throughout the rise of the novel. If we are meant to use the novel metadata file on the Github, I think we might be able to use topic modeling to explore these relationships. Obviously, topic modeling would not match specific years with narrative form or something like that, even though looking at specific dates, perhaps on a line graph, might be the easiest way to examine these shifts. Instead, I think you could do a topic modeling exercise where you use that metadata file and look at matchups between narrative form and the types of titles/words in titles. My research question would essentially be: “how can we look at shifts in narration and novels’ consciousness of its form and fictionality in a non-numeric way?” Like I said, I think we could examine this question by topic modeling the co-occurrence of certain narrative forms and book titles. For instance, we would probably see lists that grouped epistolary, first-person novels with long titles such as “the history of so-and-so” with lots of extra details included in the title. This would be contrasted with lists of more advanced narrative techniques with less laboriously-detailed and formulaic book titles. The difference between these various sets of co-occurring topics would hopefully reveal a shift in the way novels describe and think about themselves externally, in the title, and internally, in the narration.

Final Paper Plan(s)

3 min read

Throughout the course, I have noticed a shocking number of similarities between the novels and theories that we have examined and Don Quixote, which I am reading for my Spanish class. This may not seem surprising since Don Quixote is often considered a candidate for the title of “the first novel,” but it is interesting to me because Don Quixote was first published well over a century before the first novels that we examined in this class. It is also intriguing to look at the different traditions that our novels and Don Quixote evolved out of, as well as the cultural elements that influenced the novels. I have made a lot of interesting connections between our novels and Don Quixote throughout the class, but I have narrowed my ideas for the final paper down to two potential options. My first idea would be comparing and connecting Don Quixote to Northanger Abbey. They are an unlikely match, since Don Quixote is a legendary work in its own right that essentially made Cervantes, whereas Northanger Abbey is a lesser-known work by a famous writer. Additionally, Don Quixote was published more than two hundred years prior to Northanger Abbey, and there are obvious gender differences (authors and characters) as well. Nonetheless, I noticed a lot of striking similarities between the two novels which I think would be really interesting to explore, including: • Both novels parody the conventions of their genre • Yet at the same time they make a case for the novel o And they are both better versions of the previous novels that they are lambasting. In mocking the genre, they are simultaneously improving it. • Both protagonists are not naturally-born heroes. They are unlikely heroes who must develop into their protagonist role, which is a rupture with the previous literary traditions. • They both also establish an interesting relationship between novels and real life. Through their protagonists, they make fun of those who would try to see novels as exact representations of real life. Yet they also prove that novels still hold value without attempting to adhere directly to reality. o This is related to the debate/evolution of reading as pure entertainment or something more involved/useful. Both novels definitely deal with this theme, although I believe their messages diverge a bit. I am not 100% sure how I am going to synthesize all these connections into a clear takeaway about the impact of comparing Don Quixote and Northanger Abbey on the rise of the Anglo-American novel. However, I do think there is a lot to be said about the work that these novels do in terms of advancing the novel, tweaking and developing its existing conventions, and dealing with the task of reading. If I do not go that route, I would like to look at Don Quixote as it connects to Daisy Miller. There are fewer overlaps between these two novels, but a particularly striking connection between them is the idea of a character’s hidden subjectivity and an unreliable narrator. I think I could say a lot about the way that these elements, present in different ways in both novels, deal with subjectivity on the part of the reader and the advent of reading as an involved, interpretive task, rather than pure entertainment. Don Quixote especially goes even further than Daisy Miller in this area, so it would be interesting to look at the implications of that on the rise of the Anglo-American novel. Obviously, like my other idea, I still need to think through an exact synthesis of the impact of these connections.

Topic Modeling

3 min read

Topic: female roles; set-up: 50 topics, 1000 iterations, 20 printed life female art pleasure company public manners appearance friends love natural education opinion generally world fortune beauty emily young pretty Topic: male roles; set-up: 30 topics, 750 iterations, 15 printed people country men laws nature state man spirit religion equally natural society great generally influence topic: female appearances; set-up: 30 topics, 750 iterations, 15 printed company lady ladies fine good women person woman time beauty young gentleman agreeable fair appearance topic: female domesticity; set-up: 50 topics, 1000 iterations, 20 printed family time manner account woman part sister possibly moment degree mildmay nature present colonel naturally louisa give air delicacy means topic: female domesticity and appearance; set-up: 50 topics, 1000 iterations, 20 printed company fine ladies play pretty people dress table gentlemen face dinner appearance figure laugh good set head hair taste glass topic: female appearances as ornaments; set-up: 50 topics, 1000 iterations, 20 printed fine beauty ladies eyes round beautiful taste rich eye dress fair place fancy full air large country appearance figure picture topic: manpower; set-up: 10 topics, 500 iterations, 10 printed man great nature good world author people public history genius **also interesting about this folder was that there was only one mention of a female term, “lady,” which co-occurred with male terms Topic: female socialization; set-up: 50 topics, 1000 iterations, 5 printed company ladies young person acquaintance topic: male legal matters; set-up: 50 topics, 1000 iterations, 20 printed justice gentleman law man person affair made revenge put gave brought prison order carried knew court fellow bring common committed topic: male contributions to society; set-up: 50 topics, 1000 iterations, 20 printed pleasure world man find nature men wisdom mankind friend art pleasures happiness taught china society manner knowledge objects order rest

I was interested in lists that mentioned either only female or only male terms and their relationship to Nancy Armstrong’s argument about novels’ construction of societal gender roles. I found that a lot of these topics, when they only mentioned specifically female or male terms such as “man” or “lady,” generally also included nouns that have become associated with gender roles in 18th and 19th century society (ladies, pretty, appearance, home, etc. versus man, public, laws, state, etc.). Since the MALLET tool does not recognize the actual meaning of these words, there is no way it could group these topics together based on associations. Instead, it would only list them together based on their common co-occurrence in novels, therefore supporting Armstrong’s claim that novels create these gendered societal associations and not the other way around. I began investigating Armstrong’s argument in the context of these topics because of an initial topic list that seemed to contradict or at least complicate her claim, the list in which female and public appeared together. This seemed to suggest to me that perhaps women were represented as public rather than private figures in fiction but that real-life society relegated them to the home. However, the rest of the topics that involved “female” terms associated women with the home, or with public appearances and beauty. After these results, I looked back at that original topic list and it too appeared to be in line with Armstrong’s assertion about the novel’s influence on gender roles/norms. Even though female and public co-occurred in that intriguing list, the rest of the words in the list such as art, manners, pleasure, beauty, appearance, etc. did not suggest women as public figures as I had originally assumed, but rather women as public ornaments.

Descriptive Bibliography Ideas...

2 min read

Something that I’ve noticed that traditional bibliography does not describe about novels is the significance of the actual text itself. Of course, a bibliography is not a plot summary. With that being said, these bibliographies seem to me to give undue weight to the minutiae of the physicality of the book while devoting one small section of one line to the biggest chunk of pages: the text! This representation seems to me not only surprisingly lopsided, but perhaps even misleading since the same system of writing page number-page number is used for all the sections of the book, almost giving the appearance that sections which vastly disparate page counts actually take up the same amount of space, since the way that the traditional bibliography sets them up gives essentially equal space to aspects like the title page or a random blank page with a seal on it as it does to the 242-page text, not to mention the huge amount of space that the rest of the details of the book’s physical features occupy throughout the rest of the bibliography. I would like to do a project that corrects this imbalance and more accurately represents the relative significance and size of each of the features of the book described in the descriptive bibliography. I’m thinking of either re-creating the title page or the bibliography itself with the text changed to capture this misleading representation. For instance, I would try to do something where the “text” portion is in a really big, bold font while the blank page with the seal is in really small font. I would like to represent this more tangibly, for instance by comparing the sizes of stacks of paper between the 242 page text and the 10 page preface, but I have yet to come up with a physical representation that is both insightful and eco-friendly. If one of these plans doesn’t work out, I will probably make a movie poster. This would not only capture the actual relative importance of the text itself compared to the other aspects of the traditional bibliography, but it would also give more weight to other aspects of the book that the descriptive bibliography does not accurately represent, such as the author, the title of the book, and various elements of the title page used to draw the reader’s attention.

Descriptive Bibliography

1 min read

Summary: Pratt, Samuel Jackson. The Tutor of Truth. Vol 1. London. 1779. Title Page: THE ǀ TUTOR OF TRUTH ǀ BY ǀ THE AUTHOR OF ǀ THE ǀ PUPIL OF PLEASURE &c. &c. ǀ IN TWO VOLUMES ǀ VOL. I. ǀ However Hypocrisy may flourish for a time, even its ǀ happiest moments are clouded, and TRUTH shall at last ǀ prevail. ǀ LONDON, ǀ Printed for RICHARDSON and URQUMART, ǀ under the Royal Exchanger. ǀ M DCC LXXIX.

Collation: xiv. 225p. 12mo. a3b4B-T12U5 242 leaves.

Contents: i-title page ii-Harvard College Library seal iii-dedication iv-xiv-preface 1-242-text

Notes: Source location: Harvard University Houghton Library Pratt, Mr. (Samuel Jackson). The tutor of truth. By the author of The pupil of pleasure, &c. &c. In two volumes. ... Vol. Volume 1. London, MDCCLXXIX. [1779]. Eighteenth Century Collections Online. Gale. Swarthmore College Lib TRICO(PALCI). 9 Nov. 2015 The title quote is excerpted from page 289, volume 2 of The Tutor of Truth. On the blank page after the title page there is a stamp from Harvard College Library, dated April 17, 1923.

I have to admit that I have yet to be able to draw any solid conclusions from using word clouds. In this assignment, for instance, I felt that there was a lot of interesting material in the various fusion charts in regards to gaining insight into the narrative form of the novels in this list. I was excited by these discoveries and wanted to build upon them in Step 3 of this assignment, but in a way I felt like I was really going backwards into the realm of unsupported visual conclusions rather than the factual, numerical, and comparable takeaways I was able to draw from Fusion. Obviously, there are inherent differences between charts and word clouds and the kinds of information that they are used to provide—even kids at science fairs know to use a chart rather than a word cloud to represent their results on their tri-fold poster board. Nonetheless, I thought that for the purposes of “computational analysis” in this exercise (as with many in the past), I found the word cloud to be generally unhelpful.
With that being said, I was pretty intrigued by the visualizations provided in Fusion. Although none of them were particularly deeply revealing or completely unavailable by simply using the spreadsheet, I thought that the two charts, especially used together, provided some interesting insight into trends in narrative form throughout the 18th century. Obviously, I noted the significant spike in novels in general in 1769, followed by a decrease of 31 books the following year. I also noticed that very few novels were written prior to 1741. As I attempted to compare the two big adversaries in narrative form, at least according to our tracking of novels from Robinson Crusoe to Northanger Abbey, I was surprised to learn from the bar graph that there were more third person narratives than epistolary novels overall. The pie chart, however, reduced this discrepancy by pointing out that while third person novels made up 34.9%, epistolary novels made up 28.6% of novels published in the 18th century. What this pie chart did not represent that the bar graph was able to, however, were the dates corresponding to these two narrative forms. It was interesting to observe that in addition to the discrepancy between the two forms (although less significant when considered on a percentage basis), third person novels also appeared earlier than epistolary novels and experienced more variance in their popularity, whereas epistolary novels coincided with the general 1769 peak. These comparisons between counts, percentages, and dates were all much more revealing than the simple relative sizes of the different types of narrative forms in the word cloud.

Exercise 5

5 min read

Step 1: An interesting distinction I noticed in the novels listed in this bibliography was the difference between novels that seemed to me to fit more of the contemporary conventions of novels and those that much more resembled the style of the early novels that we have read so far in class. A few characteristics gave me the impression of this notable difference, one of the most salient being the question of descriptiveness. The novels which seemed most like Robinson Crusoe, Pamela, and Evelina were those overflowing with descriptive information in the title of the book. Many of the novels listed here had more than one title; a primary one such as a person’s name, followed by a clarifying phrase that conveyed the form of the novel (letters, memoirs, etc.) and the novel’s purpose and effect on the reader (to cultivate character, virtue, morality, etc.). The titles that seemed more modern to me left out all of this descriptive information and instead were simple, mysterious phrases which gave only the slightest indication of the focus of the book, thereby piquing the reader’s curiosity. These included titles such as Misplaced Confidence, The Pupil of Pleasure, and The Relapse. Although some of these more modern-seeming titles did appear earlier in the 1770s, the general trend away from the overly descriptive titles can be seen as a progression over time, with more of these mysterious and limited book names appearing in the 1780s than the 1770s.

Interestingly, I noticed the reverse of this distinction when it comes to authorship; that is, the authors that seemed to follow more recent conventions were those that included more information about themselves, whereas anonymity, or the complete lack of information, was a much more prominent feature of the earlier writers. I observed that some of the later authors in the 1780s even employed the technique that is used heavily today by movie directors, in which they boast of their previous famous works in order to sell their most recent production, such as one 1780 title which proclaims to be written by “the author of Liberal Opinions, Pupil of Pleasure, Shenstone Green, etc.”

Step 2: I observed many of the same characteristics that I discussed above in the digital versions of the novels and, specifically, in their title pages. A somewhat surprising observation that I made when comparing a variety of novels from the 1770s and 1780s was that in my categorization of the novels as more or less descriptive and thus more or less modern, Evelina seemed the most like Robinson Crusoe and Pamela and the least like contemporary novels. Evelina’s title page adheres to many of the overly descriptive conventions I have associated with early novels, including the form of the novel (letters), and a secondary, explanatory title (a young woman’s entrance into the world). Interestingly, Evelina was one of the later novels that I studied for this exercise (I looked at two from 1776, one from 1779 and one from 1781, in addition to Evelina) but its title page seemed the most descriptive and thus least modern to me. Even the 1776 novels, such as Emma; or the Child of Sorrow, had sparser title pages than Evelina. Notably absent from Evelina’s title page was any mention of the author, which again reinforced my perception of Evelina as a more conventionally early novel. This observation became especially apparent when I compared Evelina to The Tutor of Truth, also published in 1779. Although both novels emerged in the same year, The Tutor of Truth gave the appearance of being much more modern because the significance of the author was the second most striking information on the title page, and the author was announced in reference to his other popular novels.

Step 3: I felt about the Artemis tool about the same way that I felt when we used the Voyant tool a few weeks ago: it was more fun to play with and interesting to look at the results than actually revealing about the novels themselves. For instance, while I at first tried to draw conclusions about the form of the novels from the relative sizes in the term cluster wheel, I was confused by the repetition of some words in different sizes. At first I thought the term “history” was bigger than the term “novel,” which seemed to suggest something telling about the rise of the novel as an overtly fictitious title for a story, but there were also smaller “history” and “novel” parts of the wheel that confounded this conclusion.

The term frequency and popularity chart was also somewhat limited by the short time period that I chose (1770-1782). Additionally, the results that I saw there seemed to contradict my observations from the term clusters. In the term cluster wheel, for example, “epistolary novel” showed up as one of the biggest terms. I was surprised to see it appeared even bigger than “letters.” However, when I graphed these two words against one another, “letters” appeared consistently way more popular and frequent than “epistolary novel.” Thus, I did not feel that I could draw any significant conclusions about novels from this tool overall.

OCR Says "This is America, Speak English"

2 min read

Since I am reading Don Quixote for my Spanish class, I looked at Chapter X in which the parson’s horse is compared to Don Quixote’s horse Rosinante. I converted the PDF into ABBYY Fine Reader and Google Docs, which made for some interesting comparisons. Both OCR versions of the chapter messed with the format to varying degrees, although Google Docs had particularly crazy formatting. ABBYY, on the other hand, just had a few random indents and a couple pages in which the text ran off the page. Aside from this obvious change and various errors in mistaking s for f, the most notable error to me was both programs’ inability to process the words from Don Quixote. They both misspelled Quixote, Rosinante, and Spanish. Google Docs even spelled Spanish “Spani/%,” which seemed like a pretty extreme misspelling for a word that I did not think would be that difficult for Google Docs to recognize. This reveals an glaring language barrier to the OCR programs, which would obviously make it difficult for readers of the OCR versions to get the many foreign references in Tristram Shandy. Going off of this significant error, my question about digital copies of novels is that of language. Obviously in this class I believe we are reading mostly British English novels so this is not a real problem, but it does make me wonder if these kinds of programs have any way of identifying other languages and not completely butchering them in their digital facsimile. If they do not have any way to process novels that are not written in English, I think that shortcoming raises the question of divergence from the original copy even further. Novels in other languages would have to first be translated into English, losing a lot of their inherent qualities in the process. Then these translations would have to be processed through an OCR program that would further distance the format and many of the words from the original copy. At this point, it seems to me that OCR programs would essentially be retellings of the original, foreign novels that would preserve little more than the plot.

Exercise 3

2 min read

Among all the cool features that I played around with in Voyant, I found the simple word counter to be the most interesting in terms of the actual insights it revealed about Pamela and Richardson’s use of language in the book. While the word cloud was a fun visual representation of the most common words in the book, I did not feel that I could really make any concrete observations based on the relative sizes of the words in the cloud. I was able to note small things, including “Mrs.” being bigger than “Mr.” and “Williams” appearing much smaller than “Pamela” and “Master”, but these guesswork observations did not really lend themselves to any deeper understanding of the novel. However, with the word counter tool I could actually quantify the differences in sizes displayed by the word cloud, and a lot of my findings surprised me. In contrast to the word cloud, in which “Pamela” and “Master” seemed about equal in size, I discovered through the word counter that “Master” actually occurs in the book 127 more times than “Pamela” does. I did not expect this discrepancy in a book titled Pamela. I think this surprising gap can be partially attributed to Pamela’s first-person narration throughout much of the book in the form of her letters and her journal. I thought that another explanation might be that Richardson sometimes employs the word “master” as a verb, but when I searched for the appearance of the word “master” throughout the text I mostly only found it used in reference to Mr. B. I was also surprised to find that the word “virtue” only occurs 82 times throughout the text; I would have expected the count of “virtue” to be much higher since the bulk of the book is a battle for Pamela’s virtue and because Pamela holds is it as a central part of her identity. More than the word cloud or the trends graph, I felt that the text counter gave me concrete, quantifiable data about the language in Pamela. Some of the surprising results that the Voyant text counter turned up caused me to think about the way that Richardson conveys importance and gives the appearance of word frequency in regards to words that do not actually appear that commonly throughout Pamela.

Errors and (In) Accuracies
Thankfully, this time around I only had a small amount of difficulty in convincing my computer to follow the directions of the exercise. Once I surpassed that small challenge, I found that the most interesting thing about mapping the results of the NER locations list was the differences in the errors produced by the NER and by Google My Maps. Both programs made their fair share of mistakes in identifying the locations in Robinson Crusoe, but their errors were not of the same nature. As we noted in class, the NER made plenty of mistakes based upon the kind of entities that it considered locations, such as Providence or the Bible. While Google My Maps caught these mistakes and included them in the list of locations that could not be geocoded, it still slipped up in correctly identifying the context of the locations. Google My Maps revealed an obvious US-centric bias, designating Trinidad as a city in Colorado or Amazon as a town in Montana. It intrigued me that the NER committed errors in entity while Google My Maps committed errors in context—a mistake that converged in the case of Providence, Rhode Island.
This notion of Google My Maps’ incorrect preference for US locations was also interesting compared to the World Map of Cruso(e)’s voyages. On the one hand, the Google version is obviously a more accurate map in that it accounts for the hundreds of years of geographical revisions that have occurred since the 19th century. However, in regards to the actual novel, the older map more accurately represented Crusoe’s voyages by tracking his travel routes, instead of pinning down every location mentioned in the book as if Crusoe had visited each place, and by illustrating the Caribbean and European focus of the actual setting of Robinson Crusoe. Like the contrast between the NER and Google My Maps, the contrast between the inaccuracies of these two maps revealed errors in geographical accuracy on the one hand and literary accuracy on the other.

After my long and complicated battle with my PC, I am now finally able to appreciate the interest that the rest of the class has expressed for these lists. The extensive list of locations struck me as the most thought-provoking. Offhand, it seems appropriate that a novel about seafaring adventures would be brimming with references to exciting, far-off places, as the Location list suggests. The list runs the gamut from somewhat everyday locations such as England, Europe, and York, to iconic cities such as Madrid, Jerusalem, and Paris, to a variety of exotic and wild islands and "New World" destinations. However, in spite of this impressive enumeration of foreign destinations, the majority of the novel takes place in a single location. This observation relates to something I have noticed throughout the book: whenever Crusoe ventures to other parts of the island besides his "castle," he describes his endeavors as "going abroad." The common connotation of "going abroad" involves someone traveling to a foreign country, but when Crusoe goes abroad he remains on the same small island, never wandering beyond walking distance from his settlement. The diverse list and the phrase “going abroad” create a contrast between the idea of exotic adventures and the reality of Crusoe's extremely restrained, stationary existence. This disparity is particularly interesting in the context of Robinson Crusoe as a novel that was supposedly meant to show middle class English citizens to a life of extravagant travel and adventure. In spite of this conception, Robinson Crusoe focuses more on the story of one middle class man's incredibly mundane subsistence in a single place for 28 largely monotonous years. Rather than a revolution of middle class life, this seems to me to be an example of those in the middle class being unable to escape their mediocre station in life even in circumstances as extreme as Robinson Crusoe's. The two opposing stories told by the Location list and the single location in which most of the novel is set reveal a contradiction in Robinson Crusoe as an exciting adventure novel and as a somewhat uneventful account of middle class life.

p. 31

1 min read

"[a]bused prosperity is oftentimes made the very means of our greatest adversity..."