University of New Mexico researchers Teresa Neely, left, and Monica Etsitty Dorame, right

Photo credit: Rachel Whitt

Delores Tucker is often remembered for her criticism of “gangsta rap,” but she can also be credited with prompting a new form of hip-hop scholarship. In 1997 the activist and politician used several Tupac Shakur lyrics to issue a multimillion-dollar lawsuit against the artist’s estate. Teresa Neely, then a doctoral student, heard the news and recognized the lyrics as being taken out of context. To her it was a sign that Shakur’s words needed to be studied as a whole to be understood. Nearly 20 years later she found an opportunity to not only explore Tupac’s work, but also connect the late rapper’s music to another generation of hip-hop artists.

In August 2015, Neely, now the assessment librarian at the University of New Mexico’s College of University Libraries and Learning Sciences, Albuquerque, was talking with her colleague, library operations manager Monica Dorame, about the widespread media coverage of the Black Lives Matter movement and violence against black and Native American communities. Their conversation initially dwelled on the question of media coverage, but soon pivoted to music. Neely saw parallels between current Native American issues and Tupac’s subject matter from two decades ago. Dorame wondered whether contemporary Native American hip-hop artists explore similar topics. They soon realized they could study emcees’ lyrics instead of media coverage. “The lyrics,” Neely told LJ, “would have been another way to get what was happening in these communities out [to the public].” Neely and Dorame, along with Neely’s mentees, education librarian Sarah Kostelecky and learning services librarian Jorge Ricardo López-McKnight, decided to explore this question together.

While Neely was deeply familiar with Tupac’s music, most of the librarians were not particularly knowledgeable about Native American hip-hop. Kostelecky, who is Zuni Pueblo, knew of some emcees through her past work at the Institute of American Indian Art, so she recommended several artists. The group decided to select artists with music widely accessible on iTunes, so songs could be purchased and transcribed. They agreed on three artists that represented a spectrum of Native American emcees: Frank Waln, Nataani Nez Means, and Litefoot. Waln (Sicangu Lakota) and Means (Oglala Lakota, Navajo, and Omaha) are in their 20s and have received exposure through an MTV documentary series called Rebel Music (Native America: 7th Generation Rises episode). Gary “Litefoot” Paul Davis (Cherokee) is a veteran musician, actor, and businessman who has been releasing music since the early 1990s. All three have reputations for rapping about history and identity through a Native American perspective.

Focus on metadata

The librarians were eager to dig into the lyrics. However, copyright law constraints came into play. Neely had previously tried to publish an analysis of Tupac’s lyrics, but had difficulty getting clearance from rights-holders. So, when the librarians weighed how to create a publishable analysis of these lyrics, they found their solution in analyzing metadata about the lyrics instead.

Like many subjects in the humanities, hip-hop is regularly researched as: historical art form, sociological phenomenon, political movement, and aesthetic expression, to name a few. Institutions including Harvard, Tulane, and William & Mary maintain archives dedicated to aspects of the culture. Some have a specific focus, like the personal collection of Tupac Shakur materials housed at the Robert W. Woodruff Library in Atlanta; Neely used this collection for her pre-data analysis research. The Hip Hop Collection at Cornell University, Ithaca, NY, has a broader range, including event flyers, photographs, and media from hip-hop’s early years. Assistant curator Ben Ortiz fields questions from a range of patrons, from undergraduate students writing papers on hip-hop artists to filmmakers researching set designs. “Frankly, we get a lot of people who are interested in seeing the materials casually without a serious research topic or project,” Ortiz told LJ. “Somebody who’s a fan or a serious hip-hop head who wants to come look through the material for their own enjoyment.” In all of these cases, researchers investigate hip-hop through a familiar process of studying practitioners and artifacts, and engaging directly with primary sources. Fewer researchers have applied data analysis to hip-hop.

Engaging with the lyrics

Before the librarians could begin analyzing the artists’ works, they first had to prepare the data set. They found Tupac’s lyrics online, but frequently found incorrect transcriptions that required cleanup. Written lyrics to one of Litefoot’s albums were provided by the artist, but the remaining songs from his albums, along with all of Means’s and Waln's songs, lacked similar documentation. The transcription process for the Native American artists posed particular challenges, because of the librarians’ unfamiliarity with some of the language and slang. The entire group weighed in on words or phrases that couldn’t be understood. In one instance, a chance encounter with a patron from Waln’s same reservation of birth was consulted. “I went and got the lyrics and tried to ask [the patron],” recalls Dorame. “[The patron] confirmed some of the words that [Waln] was saying in his lyrics… That was good!” By the time the librarians presented their findings a year later, they had transcribed over 147,000 words and more than 200 songs. The process, the group agrees, was “brutal.” However, as time-consuming as the transcription process was, the text analysis also presented its share of challenges.

Lyric analysis in progress.

Photo credit: Rachel Whitt

The librarians’ search for text analysis software led them to a practical and popular program, Linguistic Inquiry and Word Count (LIWC). The application uses a dictionary of categorized terms to analyze bodies of text and determine the psychological value of words. The librarians found several examples of research that used this application to analyze the lyrics of popular musicians such as Bob Dylan and Taylor Swift. Seemingly easier to learn and less costly than popular applications like NVivo and ATLAS.ti, the librarians opted to use the current version, LIWC2015.

The librarians quickly learned the central importance of LIWC2015’s dictionary. As the application analyzes a file, each word is checked against the software’s dictionary. When a match is found, its corresponding category is logged. The categories include metrics (word count, words per sentence, percent of words found in the dictionary), grammar (verbs, adjectives, articles, prepositions), and a range of “psychological processes” that attempt to quantify instances of emotive qualities, e.g. affect, cognition, perception, biology, or motivation. As such, the dictionary plays a central role in the LIWC2015’s analytical ability. Between the 2007 and 2015 edition, the application’s developers factored in a range of sources—such as social media, journalism, literature, and speech—to refine the list of words. The dictionary grew from 4,500 to 6,500 words and added or significantly revised 20 categories. (Note that the LIWC does not define individual words, but looks at combinations of frequently occurring words and maps them against categories of meaning—therefore it contains considerably fewer words than a standard dictionary.) In spite of this rigorous process, the librarians found gaps in the LIWC2015’s dictionary. When they tested the program by processing a smaller set of lyrics, they found around 15 percent of the words were not recognized.

To fill in this gap the librarians developed the DKL-MN2016, a 104-category dictionary that captures words not included in LIWC2015. The dictionary adds a range of terms and expressions, from historical figures (Crazy Horse, Lincoln) to slang (steelo, jimmy, locked down). The additional vocabulary expands on existing categories. For example, the LIWC2015 category for “death” includes, “bury,” “coffin,” and “kill,” but the DKL-MN2016 draws on the emcees’ lyrics to add, “black dove,” “chalk line,” and “dances in his dreams.” The librarians’ dictionary also provides additional categories of meaning for existing words. So, the term “red road,” which could be coded in terms of color and a means of transportation, can now reference the Native American alcohol recovery program, or an idea about Native philosophy. In this manner, the DKL-MN2016 factors more voices into its analysis.

Further study ahead

The team’s research reached two conclusions. First, they confirmed that these male Native American emcees discussed socioeconomic issues in their music. Second they found that Tupac and these artists shared four relevant subject areas: colonized aggression, structural racism, oppression, and the home. The emcees took similar paths to address how these issues impact their communities. For example, Tupac frequently cited Latasha Harlins, an African American teenager whose shooting death by a Korean grocery store owner Los Angeles in 1991 was presented as both a cautionary tale and to protest racial profiling. Similarly, Waln called out the Keystone XL Pipeline as an assault on the environment and Native sovereignty. To date, the librarians have presented this work twice: in March 2016 at the Indigenous Intervention Into the “Indigenous Narrative” Conference, then in August 2016 at the National Diversity in Libraries Conference. They were pleased with the reception at both conferences, and Litefoot and Waln shared their support.

As the team has not yet completely analyzed every song, the project is still in progress. Next steps include the publication of a peer-reviewed article describing their research process and creation of the dictionary. They are also interested in expanding their scope to analyze the lyrics of female Native American emcees. Even the existing data output could undergo its own qualitative analysis. Kostelecky points out that each category can easily be further deconstructed.

The librarians feel confident the DKL-MN2016 can accommodate the lyrics of not just other emcees, but of artists in other genres. Neely tested a small sample of Beyoncé lyrics and found that few modifications were required. The librarians plan to submit their dictionary to LIWC to make it accessible to others. Though there is seemingly no end to the work, perhaps there is a third takeaway of this research: the importance and relevance of underrepresented artists working in one of the most popular contemporary mediums. “The Beatles and Bob Dylan are awesome, but we have other voices saying things just as important,” said Neely.