Methodology
Semester 1: Fall 2015
As discussed in the About Page and the Conclusion Page, our team's mission was to create a new, user friendly webpage that described how Dickinson may have wanted different variations of words, or even total lines, within the eleven poems of Fascicle 16. (To view the original website by Michele Ierardi, click here. To see Dr. Elisa Beshero-Bondar's rendition of the website, click here.) However, while analyzing these variations, our team decided to dig a little deeper into Dickinson's poems of Fascicle 16 as opposed to the different published versions of the poems. In order to do this and analyze this new data, our team had to go through many different processes and obstacles to produce what you are seeing today. Below is a breakdown of everything our team has had to do in the past semester.
To begin the process to a better website, Brooke Stewart and Nicole had to go back through the original scans of Dickinson's Fascicle 16 and re-transcribe them. After we did this, we compared them to the code that was already marking up all of the poems. (You can see the original markup here.) After fixing the transcription errors and the markup, our team decided it would be interesting to compare the original Dickinson Fascicle 16 poems to the various published versions of the poems. However, this required a new and difficult markup. First, we had to find a way to distinguish which poem line, word, or punctuation variations belonged to which published version of the poems. To do this, Brooke Stewart and Brooke Lawrence got to work on creating new witness attributes within the reading elements to define the different publications (#fh for Final Harvest, etc.). Nicole created a list of witnesses in the top parts of the XML documents so we could reference the witness attributes. To see the new markup we created, click here.
As the Brooke's and Nicole got the markup finished for the poems, Alex was working to create an XSLT to turn the XML versions of the poems into HTML versions that are readable online. You can see his XSLT here. Alex was also working on a way that would help the viewers see the differences in the variations within the original Dickinson poems themselves as well as in the various published versions. He did this by creating color coordinating buttons, which, when clicked, would show and highlight the different variations within the poems. He did this by using a CSS stylesheet and Javascript.
Next, our team definitely wanted the viewers to be able to see the original, scanned images of all of the poems, the Dickinson originals and the published versions, but we were unsure how we wanted to set this up. Dr. Beshero-Bondar recommended that we try out a new concept for our team: image mapping. What image mapping would do to the images would be to highlight the differences in variants and create a mouse-over effect where the original Dickinson line would appear. There were many steps to complete in order for this part of the project to come together, such as photoshopping the images, uploading these photoshopped images into an online generator that would allow us to create the code for the image maps, and then modifying the code to present the information that we wanted to show on the mouse-over. First, Nicole went through and made copies of all the images, so we could have the original image as well as a photoshopped copy, which shows the variants within a colored box that was coordinated with the color used for the highlighting of the variants. You can see all of the images that we created and used for this project here. Alex photoshopped the images that were originally two pages long into one image and Nicole helped by photoshopping the colored boxes around the variants. Next, Alex put all of the images into the image map generator to output the code we needed in the HTML documents. Nicole then input the text we wanted to show on the mouse-over into the alternative attributes and the title attributes. You can see our discussion about image mapping here and an example of our final HTML documents with the image mapping code here (scroll to the bottom of the document).
After we got most of the hard work finished, it was time to decide what kind of data we wanted to analyze within all of the different versions of the poems. Originally, we only wanted to analyze why Dickinson may have wanted different words or lines in her poems and how this would create a different meaning within the entire poem. However, after finishing all of our previous work, we decided to analyze how Dickinson used dashes in her original poems compared to whether the published poems kept these dashes, took them out, or replaced them with another type of punctuation. With the help of Dr. Beshero-Bondar, Nicole and Brooke Stewart created an SVG graph using an XSLT to run over the collection of our XML poems that looks at the dash reduction of the published poems compared to Dickinson's originals. This, though, took a lot of thought and searching for how to find a dash regex character in the poems. We decided to use the analyze-string function in doing this. You can see our brainstorming here. After finally figuring out how to calculate our percentages, we finally output an SVG graph that you can find here with Brooke Stewart's dash analysis.
In the meantime, Brooke Lawrence was busy typing up our Home, About, and Conclusion pages, which as you can see, was a lot of work in and of itself. After all of the hard work our team put into this project, we hope it helps you understand the poems of Emily Dickinson's Fascicle 16 better. If you have any questions or concerns regarding our analysis or methodology, you can do so by contacting us here.
Semester 2: Spring 2016
As mentioned in the About Page, the Dickinson team for the Spring semester of 2016 (Brooke Stewart and Nicole Lottig) decided to expand the site to include another one of Emily Dickinson's fascicles, Fascicle 6. We decided on Fascicle 6 because it was less about death and contained more poems than Fascicle 16 did, and we wanted a challenge. Taking on this new fascicle meant starting from scratch: transcribing and perfecting new text encoding in the TEI standard. We had many obstacles to overcome when creating the Fascicle 6 part of the Emily Dickinson website that we will describe in detail below.
We first started with a very important component that our previous Dickinson XMLs did not have: a Schematron. A Schematron acts as a handwritten manual that, when applied at the top of an XML document, tells certain rules that must be followed in the XML. We created rules such as making sure every apparatus (app) element has more than one reading element (rdg), every reading element has a witness attribute (@wit), the witness attribute values match the witness IDs (@xml:id) in our outside complete list of witnesses, and the count of all witness atribute values within a single poem equals the count of distinct witness attribute values in the poem, so there will not be any repeating witnesses except if Dickinson has a variation in her own poem. We then applied this Schematron to all of the poems in Fascicle 16 and, after the below text encoding work was finishes, to Fascicle 6. You can see our Schematron here.
To start our work on Dickinson's Fascicle 6, we had to first locate the a list of the poems that were in the fascicle. To do this, we used this website. After figuring out which poems we needed, we headed to the Emily Dickinson Archive. In order to find the images for each poem, we had to search for the poems by entering in the first line of each one. We then had to head to the library to find all of the previous publications that were used in the Fascicle 16 part of the site and a few more that also contained some of the poems that we could get our hands on from our Pitt-Greensburg library. You can view all of the sources we used on our Bibliography Page. After finding all of the images of the original Dickinson manuscript and scanning in the various published versions of the poems, we began the long process of transcribing and encoding our text in TEI (we did these both at the same time to make things a bit easier and faster). You can view our initial XML text encoding here. We also created another list of witnesses for this part of the site to have our own reference to go off of before we created our bibliography. We inserted each witness that had a variation of the poem into the top of each poem.
After we finished all of our transcriptions and coding, Nicole started to work with the XSLT. We were initially using the XSLT that Alex had made the previous semester. However, this turned out to be a major problem. The way that Alex used the XSLT to create HTML pages before worked for Fascicle 16, but because we had more variations per line among the publications in Fascicle 6, the XSLT was proving to be a huge issue. After much help from Dr. Elisa Beshero-Bondar, Helena Sabel, and Dr. David Birnbaum, many hard-core sessions of figuring out new code, and a few weeks perfecting the XSLT, Nicole had the output she was looking for (You can see our discussion about the XSLTs here.). Our Digital Humanities professors gave us feedback at the end of the Fall 2015 semester of what they mentioned could be done with how the variants are displayed on the webpage. Dr. Birnbaum suggested that we create a table row for each line and that the table cells would serve to hold the variants, which would then be highlighted in their assigned colors, and the text, so each variant would be displayed in its own line above and below others rather than side-by-side like we had originally applied. This was a praticularly tough challenge with Fascicle 16 because of all the different variations that Dickinson wrote into her own poems, and also created issues in our JavaScript toggling systems and our CSS highlighting. After much patience, perseverance, and work, the XSLTs for both Fascicles were finished and ran over our poem XMLs to create beautiful HTMLs! You can view all of our XSLT experiments, our final XSLT for Fascicle 6, our final XSLT for Fascicle 16, and an example of a raw view of what our HTML pages look like on our Github for Fascicle 6 and Fascicle 16.
While sorting all of these issues out, Nicole photoshopped all of the images to create maps of them with the variants boxed in by the publications' coordinating colors, and Brooke then went through and created all of the image-mapping over these images, so that you can see the original variant that Dickinson wrote. Brooke uploaded the images to an online generator that produced the coordinates and html code for the image-mapping (This generator is very easy to use and convienient. I highly recommend it!). You can see all of Brooke's image-mapping here. We then pasted each image-map into its correct poem at the bottom of the HTML. We now had functioning HTML pages for all of the poems in Fascicle 6!
We next began on our new task of creating an SVG graph like is present in Fascicle 16, which analyzes the reduction of dashes within the outside publications in each poem. Luckily we were able to use the same XSLT to create this SVG with a few changes in intervals, and the addition and subtraction of colored lines and circles on the graph and in the legend. You can see our XSLT here and our finished SVG graph here.
Our next task was new for us because it required new skills that we learned during the Spring of 2016. We had to create network analyses for both fascicles. Our main hardship was: What are we going to analyze?. With help from Dr. Beshero-Bondar, we decided to analyze how variants are shared among works within the fascicles. We first did the network for Fascicle 16. To begin, we used a new tool we learned about, XQuery, to create a list of source nodes (specific publications), target nodes (different publications that were not the source node), edges (the variants that connect the source node to the target node), and edges attributes (the poem title, etc.). Here is the XQuery we used to create output, and here is our output from XQuery. We then took this output and imported it into Cytoscape. We played with the colors and what we really wanted to see to create an SVG output that satisfied and created the network we wanted. You can see the explanation and our network analysis here. You can also see different network analyses we played with here. We did the exact same thing with Fascicle 6 for a slightly different outcome. You can see our XQuery here, our output here, and our finished network and explantion here.
Brooke was constantly working on the write-ups for most everything, and Nicole wrote this Methodology. She then created the layout of the site you can navigate now, adding an initial home section that encompasses both fascicles, and sections of the site for each fascicle and its related HTML pages. After all of this hard work Brooke and Nicole put into this representation of Emily Dickinson's Fascicle 6 and Fascicle 16, we hope you find our analyses intriguing and helpful in understanding Dickinson's fascicles in a new way. If you have any questions regarding anything, please do not hesitate to contact us. Thank you again for everyone who has helped us on our site and to everyone who uses and supports it!