Exploration of Portrayals of Romantic Relationships between Bilbo Baggins and Thorin Oakenshield in Fanfiction

Bagginshield: An Exploration of Portrayals of Romantic
Relationships between Bilbo Baggins and Thorin
Oakenshield in Fanfiction
Sarah Coates
Oklahoma State University Libraries
sarah.coates@okstate.edu
Link to this presentation: http://bit.ly/2pq04Kx

Overview
What are the questions being answered?
Process of answering questions
How are these questions being answered?
Results and conclusions so far
Next steps

What is being answered in this project
Has the portrayal of Thorin, Bilbo, and their relationship (whether platonic or romantic)
changed in fanfiction as the movies were released? What common tropes are being used?
What can a shift in this portrayal tell us about perceptions of romance or friendship between
these characters?
How does the fanfic writers’ description of a relationship between Bilbo and Thorin compare
to the relationship between Kili and Tauriel in these same stories, if the relationship is
mentioned in their stories?
What can the difference in how romance or friendship is portrayed in these stories to what
was shown on-screen show us about the writers’ (and readers’) thoughts on relationships?

Process
● Need to create a corpus from ArchiveofOurOwn (Ao3), but first, need to determine
criteria for selecting texts to use as the corpus and how many
○ Chose to sort works in the Bilbo Baggins/Thorin Oakenshield relationship tag by
Hits, which shows popularity
○ To get a good sample size, I chose to use the first 100 stories, sorted by hits
● Didn’t want to code or write scripts, so found a solution: Calibre E-book Management
(https://calibre-ebook.com/). With Calibre, you can download plugins and create text
(.txt) and other files.
● Used the FanficFare plugin and used the URL of the results page in Ao3 to download
and convert the stories to plain text files.

Gathering metadata
Wanted the metadata about each fic (authors, word count, tag use, etc.)
Found a tool created by Flaymebird that allows you to download the contents of
work listings into a CSV file. The bookmarklet can be found here:
http://bit.ly/2obeFMJ
I acquired all of the metadata for the 100 fics I downloaded and set up an Excel file
for later analysis of tag use to see what it can say about relationships within the story

Looking at the actual corpus
Used Voyant (http://voyant-tools.org) to analyze the texts, as it provides a lot of
useful graphs, charts, and KWIC for the corpus, making it easier to see patterns
within the texts
I chose not to run texts through a text scrubber; instead, I relied on stopwords lists
included in Voyant to filter out words
I also wanted to see how the tags and author summaries would affect the results
You can also upload an entire directory of files, rather than having to upload each
individual file or create a massive single text file with all 100 texts in it

The Statistics
Corpus word count: 11,189,613 words, with 64,306 unique word forms (note that
multiple occurrences of the same word are only counted once)
After clearing out stopwords (common words like prepositions and articles), the
most common words are:
○ Bilbo (114,516 occurrences)
○ Thorin (84,415 occurrences)
○ Said (41,907 occurrences)
○ Just (28,342 occurrences)
○ Like (28,043 occurrences)

Wordcloud created through Voyant Sinclair, Stéfan and Geoffrey Rockwell. “Cirrus.” Voyant Tools. 2017. April 2, 2017.
http://voyant-tools.org.

Tag cloud created with author-created tags in Ao3
Top terms:
Alternate universe (89)
Bilbo (30)
Canon divergence (26)
Comfort (22)
Everyone lives (19)
Hurt (19)
Angst (17)
Nobody dies (17)
Slow burn (15)
Fix [Fix-it] (13)
Fluff (12)
Romance (12)
Slow build (12)
Bagginshield (11)
AU (9)
Sinclair, Stéfan and Geoffrey Rockwell. “Cirrus.” Voyant Tools.
2017. April 2, 2017. http://voyant-tools.org.

Next steps for analyzing
Did a KWIC search for various terms to see how terms were used in context. All words were
searched with wildcards so that all permutations of the word would be found (i.e. glare,
glared, glaring)
Words chosen:
● Bilbo with the word Thorin appearing within 5 words
● Love
● Like
● Glare
● Hobbit
● Dwarf

Some preliminary results
Across all 100 stories, the words Bilbo and Thorin (and the reverse, Thorin and Bilbo) appear
within 5 words of each other 27,109 times.
Bilbo and love appear within 5 words of each other 509 times
Bilbo and Thorin also glare at each other a lot; in the first 34 stories, the word glare (not
including glared, glaring, etc.) was used 46 times.
In contrast, Kili and Tauriel only appear within 5 words of each other 466 times

Next steps (or, Lessons learned)
Need to reduce size of corpus--Voyant is having trouble crunching
that many words
Perhaps shrink corpus to top 50 stories by Hits, or limit to only those
that also include the Kili/Tauriel tag or both
Getting too many hits and results to find meaningful patterns
Re-run the corpus through Voyant and see the difference in results
and analyze data

Exploration of Portrayals of Romantic Relationships between Bilbo Baggins and Thorin Oakenshield in Fanfiction

More Related Content

Similar to Exploration of Portrayals of Romantic Relationships between Bilbo Baggins and Thorin Oakenshield in Fanfiction (20)

Recently uploaded (20)

Exploration of Portrayals of Romantic Relationships between Bilbo Baggins and Thorin Oakenshield in Fanfiction