Emily Dickinson's Poems: Series #1, #2, and #3

I did the quickest of readings against Emily Dickinson's three series of poems, and, quite frankly, I did not learn a whole lot.

Apparently her poems were published postpusteously in a series of three volumes. The size of the corpus is relatively small (29,000 words), and for analysis's sake, I did my best to divide the three volumns into individual poems for a total of 456 items. See the index page for details.

One of the most common statistically computed keywords was "day". I am also always interested in love, and for some unknown reason, I wanted to learn about the use of the word "life". Thus, I asked myself, "What words are used in the same breath as 'life', 'day', and 'love'?" To address this question I simply concordanced for these words and created word clouds from the result. Apparenlty "day" is associated with "summer's", "life" is associated with "death", and "love" is associated with "life" and "thee":

I did topic modeling against the corpus to determine whether or not any themes were specific to a given volume. I modeled the corpus for eight themes and got the following results:

    labels  weights                                           features
      like  2.09551         like little day know away never life time 
     water  0.17415   water brown song ashamed behold stop shows girl 
    prayer  0.17091  prayer years cautious lip face bride dissolve ...
     worst  0.15992  worst hope service brain fled mushroom fell lift 
   triumph  0.15240     triumph envy eden wife man nights clear moved 
    memory  0.14888  memory hand beds wake roads torrid twilight gn...
   chamber  0.14681  chamber boat hid waiting maid half-past far ma...
      sand  0.14545  sand nods gained rank landlord strangers somew...

I then pivoted the model on date, and created the following visualization, and for all intents and purposes the three volumes are very very much thematically similar:

I then wondered whether or not I could identify patterns and/or anomolies by comparing the three volumes to their use of nouns or verbs. Again, no go. All three volumes use the same set of nouns and verbs:

What did I learn? Not a whole lot. I learned that Ms. Dickinson wrote many poems, but they were short. Of the poems that got published, they seem to have more similarities than differences.

Three things come to mind regarding next steps: 1) use the traditional reading process to understand the corpus, and 2) investigate ways to specifically analyze rhythm, rhyme, and meter because, after all, the content is poetry. Third, since the corpus is so small, it might be fun to download PDF versions of the content from the HathiTrust and bind the result. After all, "I have a lifetime of things to bind."

Eric Lease Morgan <emorgan@nd.edu>
Navari Family Center for Digital Scholarship
University of Notre Dame

April 26, 2023