edit in Google

Looking into Lomax’s Questionnaire for Musicians

The dataset was created by Alan Lomax, a researcher of folk music culture in the 1940s, and it consisted of questionnaires filled out by famous folk music artists from Mississippi, Tennessee and Arkansas regarding their demographic and musical information.^[1] The data includes basic information of participants’ household such as name of each member, relationship to head of household, sex, age, marital condition and education information; highly detailed socioeconomic condition of participants; and lastly a “general music questionnaire” containing information regarding to types of music they listen to, which level of musical education they received, and what instruments they play. Questions also include their favorite songs and musicians. Some forms were filled, while some were left blank. Eventually, this dataset consists of all the answers filling the entries of forms. They were collected by Lomax, stored in the Archive of American Folk-songs of the Library of Congress, which later became the American Folklife Center.

        Lomax’s study’s two major emphases, socioeconomic information and music aptitude, propels the data to include much standardized information on these topics. While his assumption remains unclear, he clearly attempted to draw a relationship between one’s economic background and one’s ability and preference regarding music. His interviewees include John Roew, Mackinley Morganfield “Muddy Waters' ', and Joe Williams; given that the ones he interviewed include many artists who became famous later, he might intentionally focus on emerging folk music artists when selecting the sample. Therefore, Lomax’s goal might be creating an image depicting the folklife music community in a way similar to a census or a poll that aims to see through the specific group from the outside.

After reviewing this document, the major theme that came to my mind was the creator of datasets, their aims, and the significance of scholars like Lomax in creating datasets using well-defined grids that transform a population into numbers and binary. Lomax’s project was different to ones we saw before during the class: instead of the state, large corporation, polling organizations, the highly detailed dataset was created in an academic context though it highly resembled a census despite that its focus was music and truly represented a portion of the population in data. In short, when datasets no longer serve as evidence for macroscopic policy such as census data, which helped deciding electoral votes and land apportionment, or business information, such as the prototype of credit-rating agencies in the 19th century, what will they do? And what is the social significance for such a dataset made by scholar Lomax?

        For Lomax’s research, it was still highly gridized and served as a force external to the local musician’s ecosystem: it broke into the cycle. Standardization implies Lomax wanted high legibility of the sample: legible to not only him but also to the audience of all shared knowledge in the Library of Congress. As James Scott mentioned in “Seeing Like a State”^[2], grids help outsiders to gain legibility of the local knowledge. Similarly, Lomax carefully designed questions in the form of grids. He prepared a printed template where participants fill in answers instead of writing short responses, leaving minimal space for ambiguity. Additionally, he asked many Yes/No and quantitative questions when inquiring about their economic condition: number of rooms, whether owning a car, whether owning radio, etc,. Overall, with such a state-like detailed investigation according to his grids, Lomax thoroughly investigated local households’ socioeconomic background and their attitude and preference towards music. Noticeably, even for questions that allow more customizable answers, Lomax’s grids remained a strong force of normalization. For instance, when asking about “what kind of music do you like best? (and why?)”, Mackinley Morganfield answered blues, because “I can play it a little bit”; also, when being asked about “what song do you think are wrong to sing”, he simply filled “None” while being asked about “what song do you want to play on your funeral”, he answered “Haven’t thought”. If this was an interview for this artist who later won Grammy awards consecutively for about 10 years, probably the facts he stated on the questionnaire would be further elaborated and heavily stamped with his own characteristics. However, when being part of the data, Morganfield’s talent was reduced to “I can play it a bit” and his reasoning and, probably, personal anecdotes regarding why he believes that no song is wrong and why he never thought about it but John Roew did, would remain a mystery. In short, grids acted as a reducer and normalizer that made an individual artist part of the background. Lomax’s grids, like any other grids, presents everything other than those unable to be measured by a simple answer.

        This surely made me think of Middletown^[3] and how a group is averaged. When being averaged, the characteristics other than those desired simple descriptions captured by the grids will be removed from the figure. In polling, a spirit of minimalism: the need to make everything fits the grids, grows. Lomax was surely interested in folk music, however, when utilizing the grids, he dissociated himself from a music-lover. Any artist on the list becomes no more than data, though some would write more and some would write less. Probably only in a way with aloofness can the picture be made. But furthermore, different to the Middletown setting where a “normal” status was desired and outliers seems unable to fit into the image and therefore ignored even by people themselves, Lomax’s archival data still aims to integrate possibilities of exceeding the grid instead of being an indifferent power to average the population: he allows the interviewee to write longer sentences even paragraphs for bettering showing the situation such as their biography as we saw on some forms. Though such a concession does not really interfere with the big picture, as Lomax’s aim was to create a visible picture of a certain community instead of averaging all musicians.

        Moreover, the grids in Lomax’s research fall into the conjunction between seeing and sharing. Lomax’s forms were not only stored in the library but also integrated into his later papers and tried his best to provide a standardized grid for the creation of such a dataset, just like publicly accessible Gallup polls. However, by contrast, Newton’s “Chymistry” table, as an early-modern dataset, was largely ambiguous and less legible to outsiders. As Foucault mentioned in “The Order of Things”^[4], data (things) in humanists’ academic settings was firstly viewed as simply dissecting various natures of “things”. In academic settings, reality was disconnected and knowledge was kept secret. Later, symbols were gradually disarmed in terms of their secrecy. Foucault emphasized that they would show “what was marked, that which did the marking, and that which made it possible to see…”: legibility not only helps the external force to penetrate but also helps to distribute such knowledge to a wider audience.

Overall, the dataset was created by Lomax as a scholar, stored in the American Folklife Center and likely to be used in academic research. Lomax was ready to make his works legible by making the covered current situation of the population legible in a way similar to other powerful external entities breaking into the cycle. After the transition from early-modern humanists to modern scholars as Foucault mentioned, scholars in the academic sphere like Lomax become part of the players that read the population and interpret them in the standardized language and serve the universal function: sharing of knowledge.

Lastly, the digitalization of the 400 forms seems a less heavy task that would take about 250 hours (if typed, as their pictures are already available and it might take less than 4 hour to photograph them) given that all data are mostly legible and the forms were clearly printed. It was never fully digitalized before but presumably collected as references in Lomax’s publications, given that he published a number of books about folk music. If digitized, it is highly likely that there will be various mappings, which are seen as visualization of all data, where elements in socio-economic condition and elements in music-related preferences or conditions are on different axissess and correspond to each other. Also, filters might be applied in order to filter the forms that answered “Yes” or “No” to certain questions or entered certain numbers in some fields, therefore the readers are able to see whether a group of participants that share the same traits are also likely to share other traits. Furthermore, they could even link the individuals who have participated in the research to their Wikipedia entries or American Folklife Center’s records, given that they were growing musicians at that time and see whether they all succeeded later. Many could eventually apply such tools to research common traits, differences, and factors influencing those American folk songs musicians who began their careers by the 1940s.

---