Quantifying Greatness

Chatsworth House, Bakewell, United Kingdom. Original public domain image from Wikimedia Commons

Type A

My Post

Earlier in the semester, we discussed how, in the vast collections of books written throughout history, the likelihood of a book being read by multiple people, let alone millions, is highly unfavorable. Therefore, for a book to be read by millions of people, it must be very great. What then qualifies as great? High quality writing? Interesting subject matter? Uniqueness? There are probably hundreds of factors that go into determining which books are great. And for hundreds of years, these factors could only be matters of debate. With the advancement of technology, however, a new way of measuring a book’s “greatness” has emerged. That is, to have a computer compare books to find what common factors are present in the great books that aren’t in the others. 

In his book Nabokov’s Favorite Word is Mauve, Ben Blatt analyzed the common belief that adverbs only diminish the quality of writing. Through the use of a computer he was able to compare hundreds of books and found statistical evidence supporting this belief. Adverbs are only one small part of a book, though, and as Blatt mentions, they are not the single deciding factor that tells us what makes a great book. They do, however, tell us that there are statistically significant commonalities across many of the great books, and that a computer is capable of finding them. Just by looking at adverbs we might not be able to tell if a book is great, but what if a computer were to analyze adverbs alongside hundreds of other features? In which case would it be able to predict the popularity of a book before it is even released? 

In the case of adverbs, the computer didn’t reveal anything that many scholars, writers, readers, and editors hadn’t already claimed. It would be interesting, however, to analyze other features to see how our notions of good writing compare with the qualities found in the books we deem great. Additionally, it would be interesting to see how these findings influence future books. For some authors, this might be seen as a tool to help them improve their writing. For others, however, such statistical representations of great writing might seem formulaic and limiting. If an author focuses too much on aligning their book with the qualities of others, it is possible that the creativity and unique voice of the author could be diminished.

There are many ways in which this form of book analysis can influence the future of books, and not only in how authors write books, but also in how AI does. While AI can’t apply creativity to books in the same way a human does, the quality of its writing might be made to mirror that of the great books, or at least be improved, if specific features of great books (like no adverbs) were analyzed and applied to its writing. Currently, the writing produced by AI is not of the same quality as that produced by humans, but with how quickly it is improving, it will be interesting to see how long it takes until an AI can write an original book of comparable quality to Hemmingway or Steinbeck. Will we then see similar issues to what we see in AI art, with AI replacing the work of some authors? Out of curiosity, I asked Chat GPT to write a 100 word story, and based on what it provided, I certainly think it has quite a ways to go to reach “great” quality. It was interesting to note though, that it did not include any -ly adverbs.

The Bookstore of Unwritten Dreams by Chat GPT

In the quiet town of Willowbrook, a mysterious bookstore appeared overnight. Its shelves held books that foretold the future. Curious, Amelia picked up one titled “The Serendipitous Symphony.” Reading, she discovered the tale of her own life—unfolding in words not yet written. Each page turned echoed her heartbeat. Overjoyed, she embraced the unwritten chapters, savoring the anticipation. The town buzzed with excitement as others found their stories. The bookstore vanished at dawn, leaving behind dreams painted in ink. Willowbrook, forever changed, embraced the unwritten with newfound wonder, living each day as a page turned in the great book of possibility.

Chat GPT

Text analysis comparisons of books significantly impact authors’ writing approaches. By evaluating linguistic patterns, stylistic elements, and thematic trends in successful works, authors gain valuable insights into reader preferences and market demands. This process enables a more deliberate crafting of narratives, characters, and genres to align with audience expectations. Authors can refine their unique voices, emphasizing what distinguishes their work while adapting to prevalent literary styles.

Text analysis not only guides authors in optimizing language use for readability and engagement but also aids in understanding and navigating specific literary traditions or cultural contexts. By identifying successful strategies employed by other writers, authors may experiment with unconventional styles and structures, fostering innovation in storytelling.

Furthermore, real-time feedback from text analysis tools allows authors to iteratively refine their work, responding dynamically to audience responses and evolving trends. This iterative process fosters a deeper connection between authors and readers, as authors tailor their content to meet the demands of a diverse audience.

While text analysis comparisons offer invaluable guidance, authors must balance data-driven insights with their creative intuition, ensuring the preservation of individuality and artistic vision. Ethical considerations, such as acknowledging potential biases in analysis tools, are crucial to maintaining a fair and informed approach to leveraging technology in the creative writing process.

4 Comments on “Quantifying Greatness

  1. This is a great post Shaylyn- honestly it gave me a lot to think about! I like how you connected it to what we talked about earlier in the semester. You ideas about how a further study of these ideas could impact writers and potentially use writing are super interesting. I can see myself trying to use less -ly adverbs just as a result of this piece and so I imagine that a professional writer would take these findings even more seriously.

  2. You make such a great point! I think further research in this area would be highly useful in the development of AI technology when it comes to analyzing literature. I think it would be difficult precisely for the reasons you mentioned regarding knowing what forms of literature are considered “great.” Would ChatGPT be able to consider those factors when pulling similar things from other “great” literature?

  3. You are so right about AI not being able to be “creative” but can replicate great works. I think it will be particularly interesting to see how tv and movie writing will be influenced by the replication of other great works through AI. Thanks for sharing your thoughts!

  4. Your approach to this blog post was very creative. I also found several of your proposed further studies interesting, like trying to predict a book’s success based on the quality of writing. I wonder how these could be adapted to other popular storytelling mediums, like comics and TV shows.

Leave a Reply

Your email address will not be published. Required fields are marked *