Crowd sourcing the Great Books

This posting describes how crowd sourcing techniques are being used to determine the “greatness” of the Great Books.

The Great Books of the Western World is a set of books authored by “dead white men” — Homer to Dostoevsky, Plato to Hegel, and Ptolemy to Darwin. [1] In 1952 each item in the set was selected because the set’s editors thought the selections significantly discussed any number of their 102 Great Ideas (art, cause, fate, government, judgement, law, medicine, physics, religion, slavery, truth, wisdom, etc.). By reading the books, comparing them with one another, and discussing them with fellow readers, a person was expected to foster their on-going liberal arts education. Think of it as “life long learning” for the 1950s.

I have devised and implemented a mathematical model for denoting the “greatness” of any book. The model is based on term frequency inverse document frequency (TFIDF). It is far from complete, nor has it been verified. In an effort to address the later, I have created the Great Books Survey. Specifically, I am asking people to vote on which books they consider greater. If the end result is similar to the output of my model, then the model may be said to represent reality.

charts The survey itself is an implementation of the Condorcet method. (“Thanks Andreas.”) First, I randomly select one of the Great Ideas. I then randomly select two of the Great Books. Finally, I ask the poll-taker to choose the “greater” of the two books based on the given Great Idea. For example, the randomly selected Great Idea may be war, and the randomly selected Great Books may be Shakespeare’s Hamlet and Plato’s Republic. I then ask, “Which is book is ‘greater’ in terms of war?” The answer is recorded and an additional question is generated. The survey is never-ending. After 100’s of thousands of votes are garnered I hope too learn which books are the greatest because they got the greatest number of votes.

Because the survey results are saved in an underlying database, it is trivial to produce immediate feedback. For example, I can instantly return which books have been voted greatest for the given idea, how the two given books compare to the given idea, a list of “your” greatest books, and a list of all books ordered by greatness. For a good time, I am also geo-locating voters’ IP addresses and placing them on a world map. (“C’mon Antartica. You’re not trying!”)

map The survey was originally announced on Tuesday, November 2 on the Code4Lib mailing list, Twitter, and Facebook. To date it has been answered 1,247 times by 125 people. Not nearly enough. So far, the top five books are:

  1. Augustine’s City Of God And Christian Doctrine
  2. Cervantes’s Don Quixote
  3. Shakespeare’s Midsummer Nights Dream
  4. Chaucers’s Canterbury Tales And Other Poems
  5. Goethe’s Faust

There are a number of challenging aspects regarding the validity of the survey. For example, many people feel unqualified to answer some of the randomly generated questions because they have not read the books. My suggestion is, “Answer the question anyway,” because given enough votes randomly answered questions will cancel themselves out. Second, the definition of “greatness” is ambiguous. It is not intended to be equated with popularity but rather the “imaginative or intellectual content” the book exemplifies. [2] Put in terms of a liberal arts education, greatness is the degree a book discusses, defines, describes, or alludes to the given idea more than the other. Third, people have suggested I keep track of how many times people answer with “I don’t know and/or neither”. This is a good idea, but I haven’t implemented it yet.

Please answer the survey 10 or more times. It will take you less than 60 seconds if you don’t think about it too hard and go with your gut reactions. There are no such things as wrong answers. Answer the survey about 100 times, and you will may get an idea of what types of “great books” interest you most.

Vote early. Vote often.

[1] Hutchins, Robert Maynard. 1952. Great books of the Western World. Chicago: Encyclopedia Britannica.

[2] Ibid. Volume 3, page 1220.

Published by

Eric Lease Morgan

Professionally speaking, I am a librarian.

7 thoughts on “Crowd sourcing the Great Books”

  1. Is there a way to ‘tag’ the great books to the great ideas, i.e. so that relevant books are juxtaposed instead of random books that have no relevance to the idea? I think this would yield better results overall and perhaps more folks would participate.

    I work in reference at a library, and I’m also an alumna of St. John’s College, which is based on the Great Books program, so I found your experiment interesting on two levels.

    Thanks much

  2. Gina, thanks for the feedback. The short answer to your question is, “Yes”, but the long answer is, “I don’t think I want to.” Please allow me to explain.

    In a previous posting I describe how I counted the number of times each of the Great Ideas appeared in each of the Great Books. [1] From that analysis it would be possible for me to objectively add tags to the books based on the “Great Ideas Coefficient”. Thus, I would be able to assign the word “love” to Shakespeare’s Sonnets and “war” to Aristophanes’s Peace. I would then be able to juxtapose works associated with any of the Great Ideas.

    But according to Hutchins, one of the characteristics of “greatness” is the ability of a work to describe and discuss a wide range of ideas, not just a few. If this is the case, then I need to juxtapose every work to every other work with every idea. Doing the random thing covers all of those options.

    Furthermore, there is absolutely nothing wrong with answering “I don’t know and/or neither”. Since I am only in interested in which work is greater than the other a non-response does not effect the tabulation. I don’t need to know what people don’t know. I only need to know what people do know or think they know.

    Finally, the purpose of the survey is to determine if there is a correlation between my Great Ideas Coefficient and people’s opinions. So far the correlation is weak but I figure I don’t have nearly enough survey data to make a real evaluation.

    Again, thanks for the feedback. The whole thing is an digital humanities investigation in a crowd sourced context.

    [1] –

  3. I would suggest that you include Ayn Rand’s Atlas Shrugged in your survey. It discusses many of the 102 Great Ideas. It was published in 1957 but remains one of the best selling books each year. Why not include a dead white woman in the competition?

  4. John Harris, thank you for the feedback.

    The primary purpose of the project is to determine whether or not “greatness” can be measured. I am using the Great Books of the Western World as a sort of control group for this study. Alas, Atlas Shrugged is not apart of the Great Books. The survey is a way of testing my mathematical model. So far the results are not conclusive.

Comments are closed.