Liam Cole Young

Pop Music Charts and the Metadata of Culture

This talk traces a pre-history of the popular music chart—from its earliest incarnation as a list of sheet music bestsellers in fin-de-siècle roadshows to its most recognizable and paradigmatic form, the Billboard Hot-100 (ca. 1958). As a solution to the linked problems of information organization and commodity circulation, charts were essential in the emergence and management of popular music as a cultural field. In spite of this, like other forms of ‘grey literature,’ they have received relatively little scholarly attention. Charts do not offer access to recorded sound but instead compile, arrange, rank, and disseminate metadata. What data to include on the charts, how to arrange it, and how frequently to update it were decisions with long-term consequences on the popular music field: charts were a key site in the struggle for artist recognition and compensation; data points like ‘previous position’ and ‘weeks on chart’ establish the frequency and duration of song circulation and audience attention; even uses of the term ‘popular’ to describe musical genre, and a new category of fan, occurred first on the charts. Focusing on these metadata functions shows how cultural knowledge and experience is premediated by interstitial forms of writing and data organization, usually unnoticed, like the chart. In processing distinctions at the heart of the popular music field, charts are an exemplar of what recent German media theory calls ‘cultural techniques.’ Conceiving of them as such throws 20th century charts into longer histories, and futures, of listing—a technique that has structured ways of knowing and acting in human societies for millennia.

Tahani Nadim

Capturing data creatures: from specimen to sequence and back

Digital sequence data in the form of so-called “barcodes” is offering novels ways for figuring species. In the production and circulation of such barcodes practices from natural history meet archival politics and genomic promises. This talk will present three data moments for querying how specific data practices construe species’ presence: a database record, a page from a museum ledger and a scientific paper about automating species identification. It will focus on the interactional dynamics that shape and the organisational conditions that bear upon “data creatures”.

Celia Lury

Synchrony: Practising distributive uncertainty in the continuous present

Gregory Bateson was notoriously cautious about making causal claims and wary of established forms of historiography, that is, predicting future effects on the grounds of past causes. In a discussion of the culture of the Iatmul (a tribe in New Guinea) he says that it, ‘like all other cultures is really an elaborate reticulum of interlocking cause and effect. The order in which such a description is arranged is necessarily arbitrary and artificial. … Throughout this analysis I shall confine myself to synchronic explanations of the phenomena, that is to say, to explanations which invoke only such other phenomena as are now present in Iatmul culture. …’ (2004: 547; our emphasis). In this paper, Sophie Day and I discuss contemporary manifestations of the synchronic in ‘big’ data through two very different cases, the use of data analytics in UK NHS breast cancer services and emerging hashtag collectivities such as #metoo. We suggest that – in practices – these ‚personalised‘ distributions of uncertainty do not merely privatise risk (although they may do so) but recompose relations of correspondence between the individual and the collective.

Nathaniel O’Grady

The Politics of Automated Security Infrastructure: Materialising Technical Imaginaries, Public-Private Hybrids and Discretionary Power

Owing to their explorations into recent technological innovations, scholars have begun to develop new conceptualisations of automation as a set of computational practices constituted not just by inscrutable algorithmically structured processes but by its development in adaptation to, and myriad effects upon, the socio-material circumstances that it is situated within and seeks to address. The paper extends these conceptual debates and elaborates upon their bearing on practices of security through interviews with the people behind LinkNYC: a new free wifi infrastructure gradually appearing across New York City. Specifically, I unpack LinkNYC’s application as a public emergency communication device that, according to its operators, has been gradually ‘automated’. The automation of technologically inflected security practices, I argue, relies on the development of organisationally situated imaginaries and their materialisation through newly introduced calculative logics and the reconfiguration of relations across a set of data platforms. The paper then outlines the political effects that follow where new understandings of automation are applied to the security-technology nexus, showing how automation redistributes authority across the public-private hybrid of organisations collectively coordinating Link infrastructure and ushers in discretionary forms of decision making beyond its oft perceived habitat within the realms of the state.

Eva Maria Nyckel

Data Practices in Digital Process Management: Quantifying, Tweaking, Governing Labour Efficiency with and through Salesforce

The practices of algorithmic registering, tracking and controlling all kinds of phenomena in life and work environments through algorithms are continuously expanding in order to make more parameters accessible. Behavioral patterns and performances of humans are being quantified and processed in order that those processes can be algorithmically controlled. This development can be empirically observed at the ubiquitous use of process management systems in all kinds of organizations (i.e. companies as well as universities and hospitals). This contribution provides a media-theoretical perspective on Salesforce, a widely used process management system, uncovering some of the epistemological implications lying in the code of Salesforce. The paper further relates data practices that are realized through the process management platform Salesforce to registration and controlling practices of work at the beginning of the 20th century through carving out continuities and discontinuities.

Emma Garnett

Sensing bodies

I draw on preliminary ethnographic research of data practices in interdisciplinary studies of air pollution involving wearable sensors. The rise of low cost sensors in air pollution science has led to new insights about personal exposure, environmental risk and public health burden. Wearable sensors address exposure uncertainty by measuring ‘what people really breathe’, yet because the data generated is highly granular new relations, obligations and responsibilities are also invoked. Based on two ethnographic case studies of personal exposure using low cost, wearable sensors, I examine how and why people are invited to wear the sensors and what kinds of bodies are made to matter. I argue that the participatory character of sensor-based research of air pollution creates occasions for bringing social and ethical modes of inquiry to questions of exposure and harm. To end, I discuss a series of data disruptions that led to more inventive ways of performing sensor data and sensing bodies.

Tommaso Venturini

Sprinting with data

Data-sprints are intensive research and data crunching workshops where participants coming from different academic and non-academic background convene physically to work together on a set of data and research questions. They have their roots in a series of organizational innovation introduced in the field of digital humanities and open source development but have been adapted to the need of scientific research in the social sciences. Their goal is to generete research outputs in a “quick and dirty” way, in situations where the contribution of a plurality of disciplines is needed as well as the collaboration with actors outside the academia. In this talk, I will discuss the conceptual bases of this original research format and provide some advice on how to convene this kind of events.

Minna Ruckenstein

Unexpected openings in data ethnography

Ethnographies of data production, use and circulation demonstrate the value of digital data – or the lack of it – for society, organizations and individuals. By doing so, ethnographic work can aid in formulating alternative metaphors and social imaginaries for data. In my talk, I discuss how the metaphor of ‘broken data’ supported the rethinking of large digital dataset uses in research. Ethnographic sensibilities took the ‘Citizen Mindscapes’ initiative that explores a Finnish-language online conversational dataset (‘Suomi24’), consisting of tens of millions online posts, to an unanticipated direction. Content moderators joined the project as collaborators, suggesting new directions for research and guiding the rethinking of the current platform logic. The initiative was pushed to a direction that continues to make visible the human and machinic forces involved in data production and use, thereby contributing to methodological experimentation and conceptual development in research and knowledge translation.

Robert Seyfert

Algorithmic Sociality. Cooperation in Autonomous Vehicles

Current debates on Autonomous Vehicles (AV) primarily focus on technologies that are currently not available (level 4 and 5 of driving automation). In this presentation I will discuss the problems arising from this discourse. I will argue that it is (1.) dangerous because it might lead to an overestimation of this technology. Indicators of such overestimation can be found in road accidents that are related to the misplaced trust by human drivers in self-driving technologies. Beyond overestimation, the discourse on autonomous vehicles also (2.) makes invisible the actual relations between drivers, vehicles and infrastructure. Far from being driverless, so called autonomous vehicles are primarily assisted and connected technologies. I will show that these technologies rely on a radical transformation of the relations between drivers, vehicles and infrastructure, and on new forms of cooperation between them.

Julia Kurz, Dmitri Presnov

Data multiple: Diagnostics, cooperation and media on a hospital ward

We study medical data in the complex information ecology of a neurosurgical ward. As a result, we analyse medical data as operating on multiple levels of diagnostics, cooperation and media. In a first step, we will describe the information ecology of the ward showing how different kinds of medical data, e.g. clinical symptoms, body-imaging data or administrative data, are interwoven with phases of the work-flow and respective cooperative practices as well as media use. We find a complex interplay of electronic and paper-based data practices that shape diagnostics and cooperation through different modes of in/accessibility. By this, we identified cooperative-diagnostic data that constitutes the basis for developing cooperative-diagnostic visualisations. In a second step we will explain our own data practices during formalizing and enhancing such “cooperative diagnostic data” for the visualization app. For that matter, the technical solution is not designed as universal application for integrating all medical data in one device. Rather we turn our attention in building an app to fit in the existing information ecology with its well-established practices.

Isabel Schwaninger

Older Adults, Trust and Robots: Reflecting on Older People’s Experiences for Digital Care

Trust can be regarded a key part of care and healthcare, particularly when desiging technology with and for older people. Trust is however multidimensional, and can be challenging to grasp in research, and to operationalize through design. In my talk, I will give an introduction to trust-related research across HCI and CSCW. I will discuss how technology is often seen as a medium for interpersonal communication or collaboration in CSCW or HCI, while recent work in HRI has put technology in the role of a trustee. While the role of robots for (future) digital care have been taken up in the field of human-robot interaction e.g. to support older people’s independence, there is however little work taking into account people’s social practices, or focus on specific sites of matters like care. I will thus raise questions for future work on trust in robots that appear specifically relevant for older people and care. To start from older adults using technology in their everyday lives, I will draw on an ongoing AAL study that involves different kinds of technologies for people to use in their homes over a longer period of time. A qualitative analysis of the data highlights a very different way that trust plays out, not so much around the technology but around people. I will reflect on challenges regarding the design of specific AAL technologies with regards to how they are understood and adopted by older people. Taking the lessons learned from this AAL study, I will move on to the design of future technologies care. Towards this, I will also discuss a card-based method we have started to explore to initiate active conversations with older people on the topic of trust in robots, and raise questions to be addressed in future work with regards to older adults, trust and digital care.

Kate Weiner, Catherine Will, Flis Henwood, Ros WIlliams

Everyday curation: Attending to data, records and recording keeping in the practices of self-monitoring

This paper is concerned with everyday data practices, considering what records people make and keep in the context of self-monitoring. Animated by critiques of critical digital/data studies, in particular its tendency to focus on data aggregations, data flows and ‚data power‘ (Ruckenstein and Dow Schull, 2017; Kennedy, 2018), the analysis unpacks the relationship between taking a measure, making a record, and reviewing records. The paper is based on an interview study involving 67 people who monitor their blood pressure or BMI/weight. Building on, and extending, a growing scholarship on everyday self-monitoring, we adopt and develop the idea of curation to consider the role of both humans and materials in the production of records. We introduce the idea of discerning work to characterise the skilful judgements people make about which readings they record, how readings are presented, and about the records they retain and those they discard. We suggest people might produce partial data, both in the sense that it embodies these judgements, and also because monitoring might be conducted intermittently. We also explore the broad set of materials involved in record keeping in our cases, including paper, spreadsheets, devices, mobile phones, apps and social media, to consider the different ways these contributed to regulating attention to self-monitoring. Our analysis also acknowledges the difficulties people have keeping track of their records, introducing the idea of dormant records to allow for disengagements and lack of intentionality. This analysis helps to return a degree of agency to those who self-monitor and expands the view of self-monitoring records beyond the networked and digital. Ultimately, we argue, it helps to add more nuance to our understanding of the relationship between self-monitoring and the accrual and flow of data.

David Ribes

The Logic of Domains

The logic of domains has become a key organizing principle for contemporary computing projects and for data science more broadly. The logic parses collectives of expertise into “domains” that are to be studied or engaged in order to inform computational advancements and/or interventions on the domains themselves. The concept of a domain is set against a proposition that there is a more general, „domain independent“ or „agnostic“ technique that can serve to intermediate the domains. This presentation contrasts instances of this discourse, organizing and techne, drawing from cases in artificial intelligence, software engineering, and science policy to illustrate three ongoing figurations of the logic as i) experimental research, ii) formalization in method and software tools, and iii) as a de facto organizing principle for science policy and technology development.

J. Berenike Herrmann

Lovely! Books. Mining literary valuation in lay readers’ reviews

Web 2.0 users regularly read and reflect on literary books, for example in the context of large social reading platforms such as lovelybooks.de. However, the practice of lay reviews has been subject of – sometimes heated – debate: Pessimists hold that lay reviews lack the sophistication of professional reviews, are driven by hedonistic values, affirmative bias, and show an absence of a deep and discerning reading engagement. Others celebrate the democratization of interpretative sovereignty and a participatory culture. However, empirical study is still pending. My talk reports on a data-driven assessment of 1.3 Mio. reviews by 54,000 platform users, exploring the (literary) values underlying their judgements. By means of digital annotation, n-gram and sentiment analysis correlated with user ratings, our text-mining aims at implementing and testing the pertinent literary valuation model by Heydebrand & Winko (1996). Results point to an emergent type of literary valuation that combines formal-aesthetic and effect-related premises.

Mathias Scharinger

Quantifying speech melody: Multimodal approaches for studying an essential aspect at the language-music interface

Language and music share many features and structures, an observation that goes back at least as far as to the Greek root of „melody“, i.e., μέλος, referring to both song and singer. Melody is one of the shared features of language and music, with speech melody providing – at first sight – a rather elusive rhetorical dimension. Nevertheless, intuitions about speakers, dialects and individual text types, referring to patterns of intonation, can be and have been related to the description of „melodious“. Here, I present a qualitative and quantitative approach for the study of speech melody, combining individual aesthetic and affective ratings with acoustic-phonetic as well as brain imaging data. From these more-dimensional relations, it will become obvious that speech melody is by no means an elusive dimension, but rather an essential feature of the language-music interface.

Thomas Weitin

An empirical gaze? Reading experiments on poems between Realism and Early Modernism

After Wolfgang Iser the gaze of the reader must be analyzed in order to understand how our perception of literary texts works. Given that theoretical background the talk discusses some data from eye-tracking experiments on two poems by C.F. Meyer and Rilke that both try to represent the very same Roman fountain. Given those two stimuli about the same historical thing the reading experiments were conducted to find out if there is any empirical evidence for different attention patterns in Realism and Early Modernity. The respective measurement and its data, of course, need to be scrutinized facing the fact that we actually don’t know how exactly a reader performs his or her attention. Fixations could be the right measure but Iser’s theory tells us that also the opposite might be true. Irritations and the back and forth of the eye movement could be the true structure of attention. Therefore the data is inevitably contingencious.

Gaia Mosconi

Three gaps in Opening Science

The Open Science (OS) agenda has potentially massive cultural, organizational and infrastructural consequences. Ambitions for OS-driven policies have proliferated, within which researchers are expected to publish their scientific data. Significant research has been devoted to studying the issues associated with managing Open Research Data. Hitherto, relatively little interest has been shown in examining the immense gap that exists between the OS grand vision and researchers’ actual data practices. This specific contribution examines research data practices before systematic attempts at curation are made. It is suggested that interdisciplinary ethnographically-driven contexts offer a perspicuous opportunity to understand the Data Curation and Research Data Management issues that can problematize uptake. I present a detailed empirical account of interdisciplinary ethnographically-driven research contexts in order to clarify critical aspects of the OS agenda and how to realize its benefits, highlighting three major gaps: between policy and practice, in knowledge, and in tool use and development.

Wolfgang Kraus

Setting up an ethnographic data archive: challenges, strategies, experiences

The notion of Open Data rests on two main assumptions. First, data reflect the real world independently of the research context in which they are collected. They can be used to verify research results and are unproblematic to reuse. Second, once funding money has been invested, data become assets that cannot be owned by the researcher. It is obvious that such assumptions cannot hold independently of disciplinary conditions. I argue that ethnographic research provides a striking example of how mistaken they can be. The Ethnographic Data Archive (EDA) that we established in Vienna is based on a different set of assumptions. (1) Ethnographic data are not collected but co-constructed in dialog with our research subjects. (2) Thus, they also belong to the research subjects and their communities. (3) Ethnographic data have an interest beyond the immediate research context because they are usually too rich to be fully exploited in the first place and because, being situated in time and space, they are historical by nature. (4) There are good reasons to make ethnographic data accessible and reusable but they should never be treated as independent of their original research context. (4) The dialogic character of ethnography and the access to personal lifeworlds raises important issues of confidentiality, privacy and reciprocity. This means that ethnographic data can never be fully opened.

N.N.

Setting up an ethnographic data archive: challenges, strategies, experiences

The notion of Open Data rests on two main assumptions. First, data reflect the real world independently of the research context in which they are collected. They can be used to verify research results and are unproblematic to reuse. Second, once funding money has been invested, data become assets that cannot be owned by the researcher. It is obvious that such assumptions cannot hold independently of disciplinary conditions. I argue that ethnographic research provides a striking example of how mistaken they can be. The Ethnographic Data Archive (EDA) that we established in Vienna is based on a different set of assumptions. (1) Ethnographic data are not collected but co-constructed in dialog with our research subjects. (2) Thus, they also belong to the research subjects and their communities. (3) Ethnographic data have an interest beyond the immediate research context because they are usually too rich to be fully exploited in the first place and because, being situated in time and space, they are historical by nature. (4) There are good reasons to make ethnographic data accessible and reusable but they should never be treated as independent of their original research context. (4) The dialogic character of ethnography and the access to personal lifeworlds raises important issues of confidentiality, privacy and reciprocity. This means that ethnographic data can never be fully opened.

Malte Ziewitz

Black Hat, White Hat: The Ethical Work of Optimization in Web Search

When measures come to matter, those measured find themselves in a precarious situation. On the one hand, they have a strong incentive to respond to measurement so as to score a favorable rating. On the other hand, too much of an adjustment runs the risk of being flagged and penalized by system operators as an attempt to ‘game the system’. So how do those in charge of helping others rank in search engine results navigate the shifting lines between legitimate and illegitimate optimization? In this talk, I draw on ethnographic fieldwork with search engine optimization (SEO) consultants in the United Kingdom to show how SEO consultants artfully arrange themselves to cope with moral ambiguities provoked and delegated by the operators of the system. Attending to such practices of ‘being ethical,’ I suggest, does not only offer opportunities for rethinking popular tropes like ‘gaming the system’, but also draws attention to often-overlooked struggles for authority at the margins of contemporary ranking schemes.

Andreas Mertgens & Patrick Sahle

The Harold Garfinkel Papers: On Digital Archives and Visual Interfaces

Physical archives are converted into digital archives under the regime of established standards and best practices of description and object digitization. In a process that seems to be striving for objectivity but is in fact dependent upon arbitrary decisions, information objects are created and usually submitted to a single hierarchical structure. Alternative views that would create different categories, descriptions and structures are usually not realized. Traditional finding aids do not allow for additional, concurrent strategies of indexing. They organize items, but conceal aspects of mediality, context, genesis and relation. As a consequence, modes of presentation and usage in digital environments are limited to an overly simple model of catalogue, search field and result list. But is it really impossible to have multiple descriptive approaches that take into account further views on the archive objects and lead to richer presentations and browsing approaches that exploit their physical, visual, relational and indexical dimensions?

Marcus Burkhardt

Open Equals Good?: Some Remarks on the Politics, Practices and Paradoxes of Openness

The quest for open research data is entangled in broader developments toward openness, i.e. the political and practical drive for openness is neither limited to the area of research nor does it just aim for data as a matter of concern. By asking for the politics, practices and paradoxes of openness the talk develops critical perspectives on the quests for Open Access, Open Science and Open Data. While openness – much like transparency and participation – today has become a goal in and for itself, the term remains radically underdetermined. Approaches to define what constitutes openness largely remain formalistic. Yet, instead of asking “what is open?” and thereby treating openness as a matter of fact, the talk argues that openness should be treated as a matter of concern.