Spreadsheets and the Violence of Forms: Tracking Organisational and Domestic Use

IntroductionWith its capacity for modelling and “what if” logic, the spreadsheet operates as a media of beginnings and possible futures. It has proved indispensable in organisational life and labour, its failures the stuff of enduring legend about the Global Financial Crisis and the excesses of Wall Street. Indeed, the “European Spreadsheet Risk Interest Group” maintains an archive devoted to cataloguing public “horror stories” of legal actions, business failure and government enquiries due to errors in spreadsheet calculations (EuSpRIG Horror Stories). One such tale of spectacular failure occurred in 2012 when a coding error was revealed in a spreadsheet formulae used by economists Kenneth Rogoff and Carmen Reinhart to argue for the implementation of harsh austerity measures following the GFC. The spreadsheet purported to demonstrate that when debt levels exceed 90% of the size of GDP then national economies cease to grow, thereby justifying the reduction of public infrastructure and services. To substantiate the argument the data mapped worldwide national levels of debt for the period 1945-2009. However, the authors mistakenly omitted key countries from their spreadsheet – those from the start of the alphabet namely Australia, Austria, Belgium, Canada and Denmark. When these countries were included, the data showed economies can actually grow despite high levels of public spending (Yglesias). As Mike Konczal explains, the “core empirical points providing the intellectual foundation for the global move to austerity in the early 2010s was based on someone accidentally not updating a row formula in Excel”.In this paper I track the history and affect of spreadsheet use across organisational and domestic settings. Expanding upon the insights of Lisa Gitelman, JoAnne Yates and Ben Kafka, who focus on paperwork’s materialities to excavate the labour of bureaucratic media, I report on the early findings of a project exploring the role of the spreadsheet in everyday life. One of the interesting threads I pursue is how the spreadsheet becomes imbricated in the contours and vicissitudes of the home managing and recording its daily practices. Although spreadsheet applications have been extensively studied in business and engineering literatures there has been scant attention paid in the fields of cultural studies, media or, surprisingly, software studies and media archaeology. In the journal Computational Culture where one expects to see finely grained analyses of the algorithms and design decisions underpinning the spreadsheet it has not turned out to be a major concern. This isn’t to say that spreadsheets are omitted from discussions exploring the materiality of digital culture particularly from those interested in institutional life and “evil media” (Fuller and Goffey) but these are often references in passing to broader arguments. As this paper shows, the pervasive nature of spreadsheet use often eclipses its central role in our lives. In order to counter its ubiquity, to make visible its cultural impact, we need to focus on the material conditions from which it emerges and the specificities which shape its use. So I begin with a brief history of the spreadsheet format framed by legal questions of intellectual property and the ways in which these regimes enable its distribution. We then see how it operates to manage risk in relation to personal data management and the Quantified Self both in the organisation and at home. Cutting across these various sites and practices is an interest in how the spreadsheet acts “violently”, how its banality and familiarity belie its ability to generate affective intensities and real material impact. What does it mean to talk of the violence of forms? By this phrase I want to bring to the surface the ways in which forms and other administrative media operate as tools of governance. The spreadsheet in particular extends Yates’s argument in Control through Communication that the history of organisational paperwork is a story of disciplinary systems inaugurated by the circulation of internal documentation. In the early 20th-century workplace, memos, forms, tables, and circulars emerged to produce new hierarchies of managerial control through the downward communication of rules and procedures and the upward flow of reports. Recording and regulating labour, this often overlooked category of communication was instrumental in the move to quantify and monitor people’s activities at work (Yates). With properties of tabulation and calculation the spreadsheet operates discursively, it constrains emotions and bodies in particular ways. Spreadsheet HistoryThe precise beginnings of the spreadsheet program are difficult to pinpoint. Of relevance here is the distinction between spreadsheets on mainframe, timesharing computers of the 1960s to 1970s and the software developed for the Personal Computer (PC) since the 1980s. Is there a direct lineage stretching from the computerised accounting programs designed by Richard Mattessich to the VisiCalc program written by Dan Bricklin and Bob Frankston in 1979? Reacting to such a chronology, Frankston states that “Mattessich is creating his own myth” elaborating that he “completely misses the point”: I don't begrudge him his work in accounting in the 60's [sic] but it had not the slightest influence on VisiCalc. It was one of many online financial programs. I worked on some systems while at Interactive Data in the 60's and 70's. But VisiCalc was not an accounting program at all, it just made it possible for people to do accounting.This distaste for accounting is echoed by Frankston’s co-creator Dan Bricklin who explains the original name for the program, “Calcu-ledger” was rejected because it carried too many connotations of bookkeeping (Bricklin, Dan Bricklin's Web Site). Instead, VisiCal, short for Visual Calculator, speaks to its genesis in a Harvard lecture theatre where sitting as a student, Bricklin imagined “if only we had a blackboard” on which one “could erase a number and write a new number in, and everything would recalculate” (Bricklin, "How"). VisiCalc is widely thought to have been the first “killer app” in that it enabled the extensive commercial success of Apple II. Questions of law play a significant role for understanding the history and use of spreadsheet programs. One of the early mainframe based software systems LANPAR (LANguage for Programming Arrays at Random) was unsuccessful in its original application for patent approval when lodged in 1970. Its developers, Rene Pardo and Remy Landau, faced twelve years of appeals before it was granted in a landmark case of 1983 only to be reversed in 1995 (Power). Also billed as the first electronic spreadsheet by its authors, LANPAR invented the “Forward Referencing & Natural Order Recalculation” algorithm. Unlike the earlier programs or some that would follow such as VisiCal, this feature allowed spreadsheet cells to automatically recalculate rather than rely on manual refresh. In other words, the program would use a “topological sort” to calculate values of spreadsheet cells that were dependent on other cells for their totals. During this period LANPAR spreadsheet software was licensed for use by a number of large companies including Bell Canada, AT&T and General Motors (Pardo). Patent law does not generally protect mathematical calculations. Since an algorithm is a set of instructions rather than a tangible invention it doesn’t meet a crucial principle of patent law that protects creations which perform specific functions. As many commentators have noted, software occupies a curious border position: while the code is simply a string a numbers it is also executable, a process that produces or invents. Software is not included within the US patent legislation and case law since the 1960s has seen intense commercial and cultural clashes unfold over what patentability means. Title 35 of the US Code grants patents to: “whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement” (USC 35). The test for patentability requires that the invention must be “novel” and “non obvious”, that is a patent won’t be granted if the product or process is already available to the public nor if its use is obvious to anyone with an “ordinary skill” in the area to which the patent pertains. Rights were exercised for computer software through other measures of intellectual property such as copyright and trademark but in the early days of software development when Pardo and Landau wanted to register their program, code was considered un-patentable. A series of decisions by the US Supreme Court during the 1980s, known as the “Patent-Eligibility Trilogy cases”, established that a software patent claim could not be dismissed “simply because it uses a mathematical formula, computer program, or digital computer” (Place). These precedents granted their patent for “Process and apparatus for converting a source program into an object program” (Pardo et al.). Sadly, their luck did not hold out. Armed with a newly minted patent, Pardo and Landau instigated an unsuccessful law suit for patent infringement in 1989 against Lotus and Microsoft who had themselves by now developed spreadsheet software. Lotus 123 was created by Mitch Kapor and Jonathan Sachs and released in January 1983 to operate on the IBM PC, its key innovations being the introduction of macros, graphical charts and database capabilities. As the program VisiCalc had achieved with Apple II, Lotus 123 dramatically increased sales of the IBM and is one of the first software programs to run a television advertising campaign (Barker). Lotus overtook spreadsheet sales of VisiCalc who were generating $12m annually, and recorded $53 million in the first year of the program launch, ensuring it dominated the spreadsheet market through the 1980s. Although Microsoft had a spreadsheet

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.