Blog

Over the past half century, the number of MFA-granting creative writing programs in the United States has increased rapidly, from 15 in 1975, to 31 by 1984, 64 by 1994, 109 by 2004, and then an astonishing 244 by 2016 (according to the Association of Writers and Writing Programs), and graduate programs are proliferating internationally as well across Europe, Asia, Africa, and Latin America. And yet the utility of the MFA, and the wisdom of pursuing one, remains uncertain. Conferring them is surely profitable for colleges and universities but receiving one does not guarantee success in the field and can frequently leave the recipient in considerable debt. Although the degree is intended to certify competence in creative writing, it actually, and more pragmatically, prepares one to teach creative writing in a field that increasingly relies on contingent labor. Indeed, if one measured the success of any given program by the number of successful writers it produced, even the best would rate poorly. As Kurt Vonnegut once proclaimed, “Virtually everyone’s going to fail.  If you ran a school of pharmacy like that it would be a scandal.”

So why do so many people pursue MFAs, and what do they do with their lives after they receive one? And what is their role, in aggregate, in the contemporary creative economy, where content is king? We hope to answer these questions by creating a database of information about recent MFA recipients large enough to generate valid statistics for quantitative analysis, and from which a selection of individual and representative stories can be extracted for qualitative analysis. The result will be a book-length portrait of a class fragment we’re calling the Creative Precariat.

In order to gather this data, we have constructed an IRB-approved online survey with questions about age, gender, race, region, nationality, and class, alongside educational and professional information about degrees conferred, jobs held, and writing published. This data will be entered into a spreadsheet from which we will generate visualizations to illustrate our quantitative analysis. As we proceed, we will select representative individuals for in-depth follow-up interviews to illustrate our qualitative analysis. The final project will integrate these analyses, starting with a statistical overview and then moving into individual chapters focusing on representative examples.  

Do you have, or have your pursued, an MFA in creative writing? If so, please fill out our survey:  https://uiowa.qualtrics.com/jfe/form/SV_eOFf1onhZEYVoJE.  It only takes twenty minutes.

 

 

 

One sign of the persistence of the Program Era into the new millennium is the increasing prominence of creative writers, both as authors and subjects, on streaming television, indicating a socioeconomic and semiotic synergy between MFA programs and streaming content, both of which continue to proliferate despite the precarity of the career opportunities they provide. One of the first prominent instances of this synergy was Lena Dunham’s breakthrough series Girls, which was both by and about a young creative writer unstably suspended between MFA and NYC, striving for success and stature in a gig economy that at best provides only tenuous opportunities and temporary successes.

As I’ve discussed elsewhere (https://post45.org/2016/05/fail-better-literary-celebrity-and-creative-nonfiction-in-the-program-era/), Girls reflects, and reflects upon, the rise of creative nonfiction as a subdiscipline, and autofiction as a subgenre, of the Program Era, and affirms the apotheosis of the Iowa Writers’ Workshop as locus classicus of the affective anxieties and aesthetic protocols of creative writing as a career choice. Less obvious as a site of such reflection is Jane the Virgin, an American remake of the Venezuelan telenovela about a 23-year-old virgin (she’s 17 in the Venezuelan original) who is accidentally artificially inseminated with the sperm of a playboy hotelier (by his lesbian sister doctor) on whom she had a crush in high school. When I began watching this show, on the insistence of my wife, I had no idea it would have anything to do with creative writing, but it does.  

But before we even got to that subplot, I had to contend with a sneaking suspicion that the premise of the show was covertly, or even not so covertly, pro-life. After being deeply moved by her deeply Catholic abuela’s conviction that a girl must maintain her virginity until marriage, Jane Villanueva (played by the thirty-eight-year-old Chicago-born Gina Alexis Rodriguez) has fallen in love with a man she hopes to marry when she is accidentally inseminated. The entire series then turns on her decision not to have an abortion, even though that would seem to be the most sensible thing for her to do, given that she’s planning to marry and lose her virginity to her true love, a detective in the Miami Police Department (played by Canadian actor Brett Jordan Dier). But Jane refuses to have an abortion, throwing the story into motion as her fiancée and the father of her baby, both breathtakingly handsome, vie for her love. Everything that happens afterward—and a lot of things happen afterwards; in true telenovela fashion each episode compresses a lifetime of dilemmas and decisions into fifty minutes—stems from her initial choice. And many of the many plots and subplots turn on each character’s responsibility for the preternaturally cute (but temperamentally tantrum-prone) child, Mateo, who is born in the final episode of the first season and grows up over the course of the rest of the series. If Jane had an abortion none of this would have happened, not so subtly suggesting a parallel between the impetus of the show and the birth of the child, and also strongly implying an underlying moral message that the obligation to care for the child bonds the characters and brings out the best in everyone.

What kept me watching was the revelation in the first episode that Jane wants to be a creative writer!  At the center of this soap operatic narrative—which proliferates the standard plot permutations of mistaken identities, mysterious criminal conspiracies, secret twins, returns from the dead, and serial breakups and reconciliations—sits a kunstlerroman about a young woman, a Latina raised on telenovelas, who wants to get an MFA in creative writing while raising a baby as a (sometimes) single mother. This Program Era plot provides an overlay of metacommentary to the entire series, as Jane struggles to write and publish a romance novel based on her life which is, of course, also the subject of the show we’re watching. On the one hand, this means that the show includes multiple scenes and subplots mapping out the contemporary literary field, including editors and authors, MFA programs and writers’ retreats, readings and reviews, all centered on the literary ambitions of our protagonist. On the other hand, it means that the show itself is “workshopped” in real time, insofar as components of the storyline, such as plot twists or character motivations, serve as loci of creative decision-making and revision. 

Indeed, I have concluded that Jane the Virgin IS a workshop, for writing and for life, and for life-writing. The romance genre from which it springs has always been about vicariously test-driving relationship skills, especially in the days when selecting a mate was a matter of marriage for life. In the contemporary era of serial monogamy these skills remain crucial, and the endless plot permutations of Jane the Virgin can essentially be understood as speculative exercises in the challenges and opportunities presented by emotional and sexual attachments. Since Jane is also a romance writer, the show doubles as a writing workshop, as she works to attain the skill necessary to render these challenges in her writing, educating her readers in turn. By bringing these challenges together the show becomes, in essence, a workshop in contemporary autofiction; it teaches us how to navigate life by writing about it, and vice versa. 

Before we begin: here is the link to my GitHub page containing the extractor script I’ll be discussing below. This script automates the process of extracting and organizing an HTRC corpus, turning the compressed archive of page files into a collection of assembled full text volumes.

This post follows on the heels of the recent announcement that HathiTrust Research Center has expanded access to its collection of digital texts, making its complete 16-million-item corpus available for computational text analysis work. Given the challenges scholars can face with gaining access to digital texts—particularly texts under copyright—this is exciting news for computational literary scholarship. The Program Era Project’s own efforts to conduct computational text analysis on contemporary literature were only made possible through an earlier collaboration with HTRC, as I have discussed before.

Given that the scope of who can gain access to HTRC texts has greatly expanded, I wanted to offer an overview of the process I went through to prepare the Program Era Project’s HTRC corpus for text analysis. Hopefully, my walkthrough of how I extracted and organized our corpus offer some ideas or assistance to future HTRC users.

Moreover, I’ll also tell you about a tool I built that automates this extraction process, one which is now freely available on my GitHub. This tool can also be used in conjunction with other python-based text analysis scripts to automate the process of collecting text data from multiple volumes at once. In a moment, I’ll go into greater detail on how to use the tool and how to pair it with your own text analysis scripts. First, however, let’s see what the process of extracting and organizing a HathiTrust corpus looks like.

Figure 1: The Original Archive and its Contents

In this image, you see the initial archive we received from HTRC, a list of volumes and a compressed file which contains a series of folders. Each folder contains additional files, which will be accessible and useful to us as soon as we extract the initial compressed archive.

Figure 2: The Original File, Extracted

Figure 3: An Individual Folder in the Extracted Archive

The above image shows a folder containing the contents of the extracted file, each volume contained in a separate folder. The second image shows contents of a single volume folder. In each volume folder, we find a ZIP file containing the full text of our volume and a JSON file which contains metadata on the volume in question. There are important things to note about both the ZIP and JSON files, things which will impact how we extract and organize full texts.

First, it’s important to recognize that within each ZIP file the full text for a volume is not collected together as a single document. Instead, each page of a full text is included as a separate .txt file. We need to reassemble these pages to turn our volumes’ pages into a complete work.

Second, the JSON file provides a great deal of information on the text in question, information we can use for research purposes and to help organize our texts. Included in the JSON files are things like author names, book titles, and unique HTRC ID numbers for each book. These will be useful for automatically naming and placing files with our extractor tool. The JSON files also often (though not always) contain information like ISBN numbers, publisher names, and publication dates. The PEP team was particularly excited to discover this metadata, as it was information we could incorporate into our database of text metrics. This new information also enabled new ways to look at text data. Publication dates, for instance, make it possible to view trends in text metrics we’ve measured over time.

Figure 4: PEP’s Extractor Tool Running on an HTRC Corpus

The above image shows my Program Era Project extractor tool working on our HTRC corpus. It uses the metadata in the JSON to automatically organize volumes by author and title. It also extracts, sorts, and assembles the individual pages in each ZIP into a complete work. While the tech-minded are more than welcome to look directly at the tool’s code, in basic terms, the steps to turn an individual volume’s compressed files into a readable work looks like this:

• The tool pulls author name, work name, and HathiTrust ID data from the associated JSON.
• It uses the author name to create a folder for that author’s name (if one doesn’t already exist).
• It creates a sub folder within the author folder named after the volume’s title.
• It opens the ZIP file and extracts all page files to the book folder.
• It sorts the page files in numerical order and creates a single text file out of them, named “full.txt,” in the book folder.
• It places a copy of the metadata file in the folder, renamed with the HTRC ID number (e. a volume with the HTRC ID 123456 would be 123456JSON)

This gives us a single folder where all the volumes, their page files, and their associated metadata can be found. It is a catch-all repository for our HTRC corpus, which is now extracted and assembled as full texts. That said, this archive be somewhat large and unwieldy if all we’re interested in is the text files themselves, particularly in cases where the metadata might not be consistent. Discrepancies in author names, for instance, can create multiple folders for a single author, an issue that can been seen in the image below and can lead to data cleaning efforts.

Luckily, our tool gathers full texts together in another way. After the pages for a volume are organized and reassembled, a copy of the full text is placed in a folder set to be a complete repository of the full texts. In this folder (I call it “collected” above), each volume is named after its HTRC ID number. For example, a volume with the HTRC ID 7891011 would be 7891011.txt. This gives us a place where we have quick access to all our text files, and each is uniquely, consistently identified.

At this point, the extractor tool has done its work. However, we took a few additional steps to organize our collection and make it easier to navigate. Looking through the list of works provided by HTRC, the PEP team found a number of false positives, works not written by workshop authors that had been inadvertently included. We wanted to remove these authors from our list. We also wanted to find a way to relabel files in our collected folder to make it easier to find specific volume or a specific author’s works.

We turned to the HTRC ID numbers to help us remove false positives and rename the files for easy navigation. The PEP team created a spreadsheet of all the volumes in our collection. On this spreadsheet, authors name entries were made consistent and all false positives removed. Each volume entry also listed the title’s HTRC ID. Because we had the unique HTRC ID, I was able to build a script that searched for that unique identifier in each file name and automatically replaced that HTRC ID name with a new naming scheme (authorname+year). Below you can see our collected archive before and immediately after I have run the script.

Using these unique HTRC IDs allowed us a way to automatically replace enter clean, consistent data for our file names. As an added benefit, it made finding and removing all false positives simple. Because the false positives are not included in our spreadsheet, their names were not corrected. Therefore, any file that retained its HTRC ID was not on our master list and should be removed from our archive.

At this point, the PEP corpus had gone from a single compressed file to a collection of over 1000 full text volumes, assembled from individual pages, organized, and ready for future work. In the particular case of the Program Era Project, I incorporated our text mining scripts into the extraction process, producing datasets for each volume we could gather together into a larger dataset (after removing false positives).

Hopefully this account gives readers curious about working with a HTRC corpus a clearer picture of what to do to get a HTRC corpus organized and readable. However, as I said at the beginning of this post, I’m also happy to note that my github page contains a freely-available version of the extractor script I built to automation the extraction process.

The tool is designed to be as user intuitive as possible. No programming knowledge should be required, and the interface is almost entirely point and click. I’lll go through the steps to run the script here, but you can also find the readme for the extractor script on my GitHub.

To begin using the tool, your first step is to extract the initial compressed archive provided by HTRC. This should give you a collection of folders, each containing a ZIP file and JSON file. If you find a folder without a zip or JSON file, remove it. It will crash the tool. If, during extraction, you find a folder crashing your tool, check it for a missing ZIP or JSON.

Figure 5: Opening the terminal

Once the initial archive is extracted, run the script. A simple way to do it can be seen above. First, right click somewhere in the folder where the script is located and select “open in terminal.” You can also navigate to the folder manually if you prefer. After you have done that, type in the following command:

python3 HTRCExtractor-Release.py

At this point, prompts will appear on the screen. They will ask you to select three folders:

• The first folder you select is the one which contains the subfolders with ZIP and JSON files.
• The second folder is where the tool will place all the page files, metadata files, and full text files, organized by author folders and work name subfolders.
• The third folder is the tool will place all the full text files (labeled by HTRC ID).

The tool should then run, letting you know as it works through the files in your archive. It will also notify you when it has completed. It lists my e-mail address should you encounter any bugs or have any questions, comments, etc…

Note: If you are working with the tool for the first time in a HTRC data capsule, you will want to run it first in maintenance mode, so that it can import any modules it needs to run. Once the initial folder select prompt appears, you can simply close the tool. You can then switch to secure mode and run the tool as I explain above.

For users who’d like to integrate their own text mining scripts into the tool, please examine the code. I’ve marked a place that allows you to insert your own scripts easily, running them on each volume you extract. Inserting text mining scripts there also allows you access to the HTRC metadata for the volume. I’ve also included some sample code to get the assembled file up and running for text-mining, though you may likely have alternative ways you’d like to work with or open the data.

To those of you interested in getting started with HTRC corpora, I hope this walkthrough and the tool provide some useful starting points. The more we increase the accessibility of computational text analysis tools and corpora (particularly corpora still in copyright) the more we all stand to benefit from greater diversity of computational analysis research projects as well as more experimentation (and, ideally, innovation) within the field of computational text analysis.

Happy mining.

Note:  This post follows up on the post of June 28 regarding my ongoing attempts to access the administrative records of the Writers’ Workshop, for which I have had to make a FOIA Public Records Request through the University of Iowa’s Transparency Office:

Public Records Request

Thanks to the generosity of Mark McGurl I have now received a redacted version of the Director’s Files from the Frank Conroy Era currently housed under restriction as part of the Iowa Writers’ Workshop Records in the Special Collections and University Archives of the UI Library.  I received 21 pdf files totaling 626 pages.  They are organized and named alphabetically starting with “A_Redacted.”  According to the Transparency Office these files represent about ¾ of the entire series I requested.  The Workshop was provided with the opportunity to withhold materials prior to the Transparency Office’s redactions.

Before I discuss what’s in the redacted records I’d like to summarize what’s in the Workshop records as a whole (their record-keeping has in fact been quite thorough), how the authority over access to these records has shifted since my FOIA request, and the state and federal laws that will determine access in the future.

The Iowa Writers’ Workshop Records consist of 13 series of varying sizes and provenance.  Here is the content description from the Library Finding Aid:

Series I: Student Coursework, consists of photocopies of students’ works arranged by semester and class section within each semester. It is the largest series in the collection, dating from Fall 1965 to the present. Note that a few of the semesters are filed out of chronological sequence.

Series II, Award Competitions, consists of writing entries from individuals vying for scholarships and other awards.

Series III, Students and Alumni, consists of files containing correspondence, applications, and other material, arranged alphabetically by name of individual within each accrual. Note that accrual dates cover academic years  e.g., 1986-91 covers the 1986-87 to 1991-92 academic years. Restricted access.

Series IV, Faculty, is arranged alphabetically by name of individual. Restricted access

Series V, Director’s Files, consists of correspondence and other material created and received by the Office of the Director. Restricted access.

Series VI, Administrative Files. Restricted access.

Series VII, Accepted  Not Coming. Restricted access

Series VIII, Rejected Applicants’ Evaluation Sheets. Restricted access

Series IX, Applicants’ Letters of Recommendation. Restricted access.

Series X, Ephemera, includes posters and other printed matter, dating from 1982 to present.

Series XI, Stephen Wilbers Project, consists of correspondence and interview notes prepared by an alumnus of the Workshop who prepared a history of the program in 1980.

Series XII, Jean Wylder Project, consists of survey responses obtained from numerous alumni during the early 1970’s as part of a history project. The responses are arranged by era of attendance/graduation.

Series XIII, Newsletters, consists of newsletters released once or twice yearly since 1970 chronicling the publishing activity of Workshop alumni and students, as well as Workshop programs and events.

The original restrictions consisted in consulting with the University Archivist and the Workshop.  Since my request, access to Series I-IX of these records comes under the authority of the Transparency Office. This includes the 380 boxes of student coursework running from 1965-2011, which prior to my request were under no restriction at all.  Anyone wishing access to these records must now file a FOIA request with the UI Transparency Office and pay for its processing (as an example, the charge for reviewing the Director’s Files was $1080; the estimate for processing the remainder of the administrative records is $12,000).

Two laws are used by the University of Iowa Transparency Office in redacting requested records, one federal and one state:

• The Family Educational Rights and Privacy Act (FERPA) (20 U.S.C. § 1232g; 34 CFR Part 99), that protects the privacy of student education records. The law applies to all schools that receive funds under an applicable program of the U.S. Department of Education.

https://www2.ed.gov/policy/gen/guid/fpco/ferpa/index.html

And

• Chapter 22.7 of the Iowa Code concerning the Examination of Public Records.

https://www.legis.iowa.gov/docs/code/22.7.pdf

I anticipate that these laws, and others like them, will restrict access to educational records for future researchers into the history of the Program Era.

 

So what’s in the records that I received and what was redacted?

Selectively and roughly alphabetically here are some examples:

A redacted copy of the AWP’s 1992 Survey of Creative Writing Programs filled out by Conroy.  The redactions are interesting, and indicate some of the difficulties researchers will face in determining the demographic composition of creative writing cohorts.  Since the law apparently dictates that any statistical measure of 6 or fewer risks exposing individual identities (I say apparently because I couldn’t find this specified in either law), we can know that in 1992 the Workshop had 55 men and 45 women enrolled, but we can’t know how many (if any) of them were American Indian, Asian, Black, Hispanic, White, or “Other.”  Nor are we allowed to know the gender or ethnic background of the faculty (though we do know that 7 are Full Professors and 5 are Adjuncts).

A_Excerpts 30

Extensive discussion with the upper administration concerning the awarding of Teaching-Writing Fellowships (TWiFs as they were colloquially known) and their importance in competing with other programs for students.  All student names in these discussions are redacted.

Much discussion with a series of English Department Chairs regarding the administrative autonomy of the Workshop.  Indeed, the process whereby the Workshop achieved this autonomy, ultimately resulting in their move to Dey House, is documented in some detail in these pages.

An itinerary for, though no results from, the External Review conducted of the Workshop by Nicholas Delbanco and Stephen Tatum in 1992.

A series of increasingly testy exchanges about violations of the smoking policy in the English-Philosophy Building (EPB).  Though the violator is never named, it is well-known that Conroy was a chain smoker.

Correspondence with or about Frederick Barthelme, Saul Bellow, Annie Dillard (about the possibility of adding a Literary Nonfiction Track to the WW), Gail Godwin, John Irving, Norman Mailer, Tom McGrath, Joyce Carol Oates, George Plimpton, and Roger Strauss.

A Graduate Writing Faculty Assistance Survey issued by the University of Houston and filled out by Conroy which confirms that faculty meet with graduate students a “fair amount” in “bookstores, local bar and local restaurant (Foxhead, The Mill),” in a “miraculously open” climate, and affirms that the criterion for acceptance includes “the pressure of the soul behind the language.”

C 15

Extensive correspondence, internal and external, some redacted, involving funding, for TAships, visiting writers, copy machines (they made 100,000 copies a month), and more.

Numerous redacted nominations for prizes and awards.

A letter to all Workshop Faculty discouraging them from conducting workshops in their homes.

 

This is only a selection; there is more in these records, far more than I have any use for (and none of it violating any privacy or confidentiality laws). Unfortunately, I’m not sure where the funding will come from to redact the remaining administrative records, nor where they would be housed if the funding is secured.  At this point, what I’ve received is housed in a folder in my Dropbox account.  Do let me know if you’d like access.

 

We’ve been combing through the index of the HathiTrust Data Capsule [more on how it was put together here].  We requested any volumes authored by a list of 380 writers and got roughly 2500 volumes back.  After deduping and weeding out false positives, we have about 1500 unique volumes representing 234 Iowa-affiliated authors, with publication dates ranging from 1913-2011*.

We’re interested in analysis on a couple of levels – first, generating StyleCard and LitMap analytics for individual authors, and eventually, exploring other trends that can be found in Workshop writing  when we look at workshop texts in aggregate [more information on the tools here].  In light of this, here are some thoughts and questions on my mind as we look over the capsule.

Multiple authors

The co-authorship problem: how do we handle texts that were co-written by multiple authors?  Many volumes are anthologies of works by several authors, sometimes limited to Workshop graduates and sometimes not.  How shall we go about extracting text by just the authors we’re interested in (without having to do it manually)?

We also have some instances of books written by one Workshop writer, with a segment by another Workshop writer – an anthology of short stories by William Kittredge with a foreword by Raymond Carver, for example.  In some contexts this might not be a problem, but when we start comparing author signals with StyleCard analytics, we don’t want to muddy a signal with text by another writer of interest.  That being said, who is including who in anthologies is an interesting question in and of itself.

Anthology duplicates

Many of the authors we’re interested in wrote primarily poems, short stories, or essays, and they’ve been published in anthologies.  Sometimes the same poem or story will appear in multiple anthologies.  How do we avoid double-dipping, and magnifying the signal of some works out of proportion to others (again, without having to do it manually)?

Unequal distribution

Some authors we were interested in, of course, are not included in the corpus at all.  Of those that are, some are much more strongly represented than others.

About half of the authors represented have only one or two volumes in the corpus. Without further scrutiny I can’t say say whether this is more due to a real difference in author prolificacy, or their representation and findability in the HT corpus (or, likely, a combination of the two).  For author-specific corpora, this is going to mean a much more reliable/well-sampled author signal for some writers than others.

For any analysis looking at some kind of “Workshop style”, we’re going to have to correct for this somehow.  How should we go about this?  Maybe chopping up works into segments and choosing random samples?  Selecting representative works for each author based on some rationale relating to literary qualities or critical reception?  Generating individual author metrics and combine them instead of running analytics on the entire corpus at once? Or, alternatively, should some authors be weighted more than others, and do we have the numbers in the corpus to do so in a way that we might want?

Derivatives

There are a few dozen translations in the corpus, the majority of them by Robert Bly (accounting for a little under half of his included works).  We will likely not include them in initial stages of analysis, but I’m very interested to see if we notice distinctive differences between his original works and his translations!

An interesting tidbit: we also have a few works in the set that are musical arrangements of poetry by Workshop writers. It seems that we have those particular texts present in other formats, so they likely won’t be included in the analysis, but it was neat detail.

---

* The Workshop was not founded until 1936.  Included publications predating the Workshop were written by early faculty, such as Edwin Ford Piper.

The screenshot above documents an exciting moment for our ongoing collaboration with the HathiTrust Research Center. It is a screenshot of my computer, remotely connected to a HathiTrust machine, running my Program Era Project text-mining tools on sample HathiTrust text data. In short, it’s a proof-of-concept, a confirmation that the PEP tools are ready to begin collecting data on thousands of texts produced by creative writers affiliated with the University of Iowa, and that the PEP team can begin to join that data with the wealth of institutional, biographical, and demographic data they have already collected.

This screenshot is a result of HathiTrust’s selection of the Program Era Project as a 2017 Advanced Collaborative Support award winner. HTRC is a collaboration between partner universities that houses an expansive digital library. As HathiTrust’s site explains, the ACS program is:

a scholarly service at the HathiTrust Research Center (HTRC) offering collaboration between external scholars and HTRC staff to solve challenging problems related to computational analysis. By working together with scholars, we facilitate computational-oriented analytical access to HathiTrust based on individual scholarly or educational need.

For the HathiTrust/PEP collaboration, the approach chosen was to establish a “Data Capsule,” a machine maintained and secured by HathiTrust, that PEP team members can remotely access and then run text mining experiments on a corpus of texts held in HathiTrust’s collections. The Data Capsule approach is crucial, as the texts to which we require access remain in copyright; they simply aren’t accessible in digital form for large-scale data collection. The Data Capsule configuration allows full texts of HathiTrust works to be measured by text mining software, but only the metrics collected by the tools can be moved off the Data Capsule machine. In PEP’s case, this means .csv spreadsheets of data on individual texts.

Now, thanks to the HathiTrust/PEP collaboration, the tools I created for the Program Era Project (described a bit more here) can be employed on a large volume of digital texts. They can be used not just for experiments, but to begin building a database of metrics on features of creative writing at the University of Iowa. For the Data Capsule collection, the Program Era Project team assembled a list of roughly 400 selected authors associated with the Writers’ Workshop and the Nonfiction Writing Program. Since receiving this list, the HathiTrust team has endeavored to find all the works held by HTRC associated with these authors. At present, over 2000 volumes have been connected to the PEP authors list. These items are then made accessible on the Data Capsule and, using the PEP tools, converted into metrics which are stored in the PEP database.

So, what data is the Program Era Project collecting? Currently, I’ve built two text mining tools for the Program Era Project. Both are written in Python and draw on Stanford’s Natural Language Toolkit (NLTK). We call them Style Card and LitMap.

Style Card is a text analysis tool that measures features of literary style such as vocabulary size, sentence length, adverb and adjective usage, and frequency of male and female pronouns. The last metric is particularly interesting, as it can provide a quick impression of gender representation trends in a work or collection of works. Additionally, by collecting the same metrics from multiple authors or multiple works from one author, stylistic comparisons can be made between aspects of later or earlier works of a single author or the complete corpora of two authors. It is, in short, like creating baseball cards for authors and literary works, snapshots of information that can be used to establish or test hypotheses.

LitMap is a software package that tracks location references in literary corpora, making it possible to analyze regional representation in literary works. This allows us to see the influence of an author’s biography on their literary output as well as measure the influence of authors migrating to and from creative writing programs on the settings of their writing. Using LitMap, we’ve already made some interesting discoveries about the frequency with which works written by authors who taught at and/or attended the University of Iowa mention the state of Iowa and the region of the midwest. The team is looking forward to sharing more with you on that topic in the future.

What’s truly significant (and truly promising) about the data these tools collect is that it will be stored in the PEP team’s database. When the ACS data is incorporated into the PEP database and available to future users of the PEP web presence, users, at a glance will be able to rank and compare features themselves.

The images below represent a proposal for the eventual look and feel of the Program Era Project web presence. The numbers used are drawn from data already collected with the PEP text mining tools. As the first figure shows, a user could rank writers by average sentence length, learning, at a glance, which authors typically create sprawling (or terse) sentences. The second image ranks authors based on the ratio of male pronouns to female pronouns. The larger the number, the more often male pronouns appear compared to female pronouns.

Users could also compare two authors—or an individual author to a control corpus—and look at differences such as first-person and third-person pronoun use (a potential indicator of narrational patterns and preferences) or adverb and adjective ratios (which can index spare or detailed prose). Scholars could see at a glance how an author’s stylistic features might compare to their advisor, how they compare to other writers in the corpus, or to a baseline corpus of writing in English.

In the following image, produced using plotly, the platform we currently use to visualize LitMap data, we see another way the PEP text tools will provide new insights into literary corpora. The image shows the strong representation of Iowa in a literary corpus comprising 75 novels by Workshop-affiliated writers, documenting how their time at Iowa has left a mark on where they write about.

The idea behind offering these metrics to future users of the Program Era Project website is that access to this information will prompt curiosity and exploration. Moreover, when users find an interesting pattern or phenomenon in the data, we hope it will prompt a direct investigation of the works included in the data. In short, beyond just presenting this information, we believe that the ability to skim over these metrics will inspire scholars to dive deeper into the texts the data is drawn from. These objectives of encouraging emergent research and driving curiosity are at the heart of Style Card and LitMap’s other principal innovation: the use of clear, easy to understand metrics. The fields of stylometry and text analysis have developed techniques that allow for astounding technical and scholarly achievements, author attribution being a notable example. However, understanding how a piece of software or a quantitative approach arrived at the conclusions it did can be difficult for users not familiar with the theoretical foundations or technologies employed. To this end, metrics tracked by Style Card were selected so that users are offered information that is easy to understand and transparent. By using simple numbers, StyleCard metrics allow any scholar—whatever their experience and training with quantitative analysis—to benefit from the Program Era Project website, broadening the number of academic projects that might be inspired by quantitative analysis.

Even better, both the Style Card and Lit Map tools were developed in such a way that anyone can use them. You simply click the program file, type in the name of the author and work you are scanning and select the name you want for the output file you will create. The tool does everything else. What this means is two-fold. First, it allows more collaboration with building our database of text metrics. If a team member can access the Data Capsule, they can easily run the software to collect metrics. Secondly, these tools will eventually be made freely available online. Therefore, any other project team that wishes to collect the same metrics will have the option available to them. Because the tools will be open source, users will also have the option to modify, adjust, and tweak the technology to their own needs. Moreover, any school that might be interested in learning more about their own history of Creative Writing, or any school that might wish to establish a partner project to the one here at the University of Iowa, will have the necessary technology.

All that said, I hope you can see why we are so excited about the screenshot of our text tools running in conjunction with the HathiTrust Data Capsule. The image represents another significant step along the way to our goal of providing students and scholars of literature the ability to explore the history of the Iowa Writers’ Workshop—and the history of Creative Writing at Iowa—in a way that was never before possible.

For folks who are interested in the backstory behind the article recently published in the Iowa City Press-Citizen, this blog post will fill you in.

http://www.press-citizen.com/story/news/2017/06/19/iowa-writers-workshop-archive-costly-search-scholar-finds/391011001/

After lengthy deliberation and extensive work in the University Archives, I’ve decided to write a literary history of Iowa City to be called City of Literature.  I have already repeatedly consulted the papers of Paul Engle, George Starbuck, and John Leggett, but material from the Conroy era was under restriction.  According to the online finding aid, this material was accessible with permission from the Writers’ Workshop. In March, I decided to request permission for access to these records.

To my surprise, I was instructed to make a request through the Transparency Office. When I wrote them to confirm they agreed that this was highly unusual; indeed they later confirmed that they had never received a comparable request and have never worked with the University Archives. Here is their portal:

https://publicrecordsrequests.iowa.uiowa.edu

After some deliberation, I decided to go ahead and submit the request. A few months later I received this cost estimate:

Professor Glass,

Below are descriptions of Series IV-VI, the approximate number of pages to be reviewed in each series, and the estimated staff time involved with reviewing each page for confidential information or other information exempt from disclosure.

This estimate includes only the time it will take to review all of the records and to identify which ones are subject to disclosure (in full, or with redactions) under the state open records law. If you decide to proceed with this request, we will send you an invoice and begin the process of reviewing the records after we receive your payment. Once the initial review has been completed, we will send you a second estimate for the time it will take to then copy or scan the records which may be disclosed and to make any necessary redactions. At this time, we are not able to provide you with the second estimate because we won’t know how many records will need to be copied/scanned or redacted until the initial review is complete.

As we have previously informed you, and as the descriptions below indicate, many of these records are likely to contain information exempt from disclosure under Iowa Code Chapter 22, such as confidential personnel information or student/FERPA records. Although we are estimating that a considerable amount of time will be required to review all of these records, please be aware that many records may still end up being withheld.

The review times below were calculated based on an average rate of 8 seconds per sheet for Series IV, and 10 seconds for each sheet for Series V and VI (due to greater complexity among the records in the latter two series).

Series IV (Faculty series)

Description: Letters of recommendation, evaluations, criticisms of submissions, applications, appointment sheets (which contain hire and separation dates, position description, date of birth, salary, SSN). Arranged alpha by name.

1 box containing approximately 3,240 sheets

Time estimate for review: 7.2 hours

Series V (Director’s files series)

Description: Administrative correspondence of Workshop director Frank Conroy, which includes correspondence regarding faculty and staff appointments, internal funding, external fundraising, personal professional development activities, and letters of recommendation

1 box containing approximately 3,240 sheets

Time estimate for review: 9 hours

Series VI (Administrative series)

Description: A wide range of records, such as financial aid records, Ida Beam Professorship files (including applications, evaluations of candidates), gift acknowledgments, correspondence including letters of recommendation, award competitions, self-studies.

21 boxes containing approximately 63,720 sheets

Time estimate for review: 177 hours

Total time estimate for initial review: 193.2 hours (billed at a rate of $30/hour, with the first hour free of charge)

Please let us know if you would like to proceed with all or part of this request, and we will send you the first invoice. We will begin reviewing the records once we receive your payment.

 

NOTICE TO RECIPIENT: THIS MESSAGE AND ANY RESPONSE TO IT MAY CONSTITUTE A PUBLIC RECORD, AND THEREFORE, MAY BE AVAILABLE UPON REQUEST IN ACCORDANCE WITH IOWA PUBLIC RECORDS LAW, IOWA CODE CHAPTER 22.

As you can see, this adds up to at least $6,000, and that doesn’t include the cost of acquiring the records I need for my project. I am currently looking into funding possibilities. In the meantime, I will be posting my progress, since this entire process is now a matter of public record and public concern. Feel free to post comments and suggestions.

More to follow!

Here’s a mockup of one visualization we’re considering to access MFA cohorts by year. As you can see, 1932 conveniently emerges as Year One of the program era insofar as the first two directors of the Workshop, Paul Engle and Wilbur Schramm, both received MAs, alongside Wallace Stegner, who would go on to launch the program at Stanford. Together, they cover the three main genres–poetry, fiction, and nonfiction–that would come to dominate creative writing programs. Norman Foerster, founder of the College of Letters and persistent advocate of the creative thesis, directed Schramm’s thesis. We’ve still haven’t identified the directors for Engle and Stegner, but Foerster is a pretty likely candidate.