The fifteen chart questions were read to learn how a question about a music chart is designed before anyone answers it, and the answer holds across them. Each asks what a chart observation shows and names the material needed to read it, and each sets beside the question a limit of its own that names one conversion the number may not undergo: into the favour of a public, a nationality or a place of residence, an absence of listeners, a rate of penetration, a measure of loyalty, a cause or a degree of contribution, a creative direction or a forecast of success. This study has called that design a refused conversion. Where the questions turn to what should follow from a chart, they set it beside cost, goals, accountability and reversibility, and they keep further inquiry and holding back open beside acting. Several questions examine the person reading the chart rather than the artist charted. Dionyas publishes these questions as editorial question designs with no judgment, and this study has read them only as designs.
Discussion
The first debate was about commensuration and quantification: what a common metric makes comparable and when people insist that some things cannot be compared. Espeland and Stevens, Commensuration as a Social Process (1998), discuss the stakes of calling something incommensurable. The fifteen questions are, on this study's reading, a published instance of such calls, with three features worth noting. They are specific: no limit says that charts mislead in general; each names one pair of things that the chart must not join, and most give a reason. They are local: the limits do not open with a domain's caution line, as the boundaries of most inquiries in the list do, but are each question's own, so that the record set as a whole holds fifteen different lines of incommensurability. And they are early: the records carry no artist, no work and no year, so that the refusal precedes the temptation. Porter's question about the prestige of quantitative methods, in Trust in Numbers (1995), bears on why such limits might be needed at all; the limits on population and on the average answer it, in this study's reading, by naming the figures that look most like objective rates and saying what they are not.
The fit is not perfect. Espeland and Stevens, in A Sociology of Quantification (2008), set out dimensions of quantification and call for an ethics of numbers; the questions offer no ethics, only refusals. And boyd and Crawford, in Critical Questions for Big Data (2012), interrogate the assumptions of large data in general, while the questions interrogate one kind of observation at a time. What the questions add to that literature is a form: a question and its refused conversion, written as a pair, so that the limit travels with the question wherever it is used. Mennicken and Espeland, What's New with Numbers? (2019), ask for quantification to be tracked across sites. The fifteen questions track it inside one site, a chart, across the many places where its meaning could slip.
The second debate was about public measures and their effects. Espeland and Sauder, Rankings and Reactivity (2007), describe how public measures change the people they measure, through self-fulfilling prophecy and through commensuration. Strathern, in 'Improving ratings': audit in the British University system (1997), describes audit that takes on a life of its own. Anand and Peterson, When Market Information Constitutes Fields (2000), show from a music chart how changes in the scope or methodology of market information can jolt how a field understands itself. This study reads the fifteen questions as written with all three in view. In this study's reading, six of the limits concern scope and method in Anand and Peterson's sense: the cut of a list, the supplier's dates, the platform's users, the countries observed, the store's territory, the lead-artist record. The two response questions concern reactivity directly. They name the two places a chart would begin to change the music, promotion and creative direction, and say that the chart alone does not decide them. Lamont's categorisation, in Toward a Comparative Sociology of Valuation and Evaluation (2012), appears in the collaboration question, where the format of the record files a shared song under one name before any valuation begins.
What the questions do not ask is as telling as what they refuse. None asks how the chart is made, by whom, or with what interests; the questions take the chart as given and limit its reading. None asks what the artist or the listeners make of the chart; the material named includes unprompted responses and original texts, but the people charted appear as subjects of observation. The response questions name the goals of the person responsible but not the goals of the artist. This study does not say these questions should have been asked. It says that a published set of chart questions has drawn its limits around the reading of the number and not around its making, and that this choice shapes what any later answer could find.
Contributions
This paper makes two contributions and one of method. To the study of commensuration and quantification it adds a published, unanswered set of questions in which incommensurability is stated record by record, as a specific refused conversion paired with each question, before any number is read. To the study of public measures it adds a reading of how a question set tries to prevent reactivity in its design: by keeping further inquiry and holding back open beside acting, by setting a chart beside cost, goals, accountability and reversibility, by stating that rank does not decide creative direction, and by turning several questions on the reader of the chart. The contribution of method is a measurement of where each record's own words lie. Across the list the boundary is among the fields most copied from record to record; in these fifteen it is each record's own, while the scope and the state are shared. A reading that carried over the usual assumption would have discarded the limits as frame and kept the scope as evidence, and would have missed the design entirely.
Limits
The evidence is fifteen records from one studio, chosen by a rule that leaves out the three chart questions linked to the foresight domain, so the reading says nothing about how Dionyas frames change in chart observations over time. The records are in Korean and are read in one translation made for this study; small choices of English wording can move emphasis, and the counts were made on the Korean to limit that. The word tests are simple, and a limit phrased without the words the script looks for would be missed. Dionyas does not say who drafted these records, and this study cannot say how the limits came to be written. Nothing here describes any real chart, platform or market, and nothing here is business, marketing or investment advice.
Further research
Three lines follow. The first is to read the three chart questions on momentum and timing with the same measurement, and see whether limits written about change over time take the same form as those written about a single reading. The second is to set these designs beside the accounts of people who read charts for a living, in labels, distributors and the press, and ask which of the refused conversions they recognise as real temptations and which they would add. The third is to follow any of the fifteen if Dionyas later links an artist, sources or a judgment to it, and see whether the limit written in advance holds once a real number is in view. This study has read these records as designs for questions about chart numbers. Its contribution is to show that each question carries, beside it, the conversion it refuses, and that the questions about response leave holding back open.