Interdisciplinary Symposium Ethnography 2.0 – Anthropology and the Digital World

  • Ethnography 2.0, symposium, University of Ljubljana, interdisciplinarity, anthropology, data, digital anthropology, data mining, data ethnography

What do terms such as data mining, big data and artificial intelligence mean to anthropologists and social scientists? And what has ethnography got to do with it? We looked into these (and other) questions at the interdisciplinary symposium Ethnography 2.0: New Approaches to Understanding Ways of Life. The event was organised on 31 May 2018 at the University of Ljubljana’s Department of Ethnology and Cultural Anthropology, co-organised by ZRC SAZU (Institute of Slovenian Ethnology) and Faculty of Computer and Information Science (Bioinformatics Laboratory), and supported by the PEOPLE project.

Anthropology and Technologies

Increasingly, anthropologists – once typically imagined as lonesome adventurers on long-term fieldwork in faraway exotic places – are dealing with the digital world and accompanying “smart machines” in high-tech fields. For instance, in interdisciplinary projects dealing with innovation in the field of emerging technology or large data sets and digital environments that are inescapably altering our ways of living and shaping our future. After all, it is people from diverse social settings and cultural contexts that are eventually using these new products and services, and anthropologists are professionally concerned with what people do, think, and say in their everyday lives. It is not surprising that the methods, such as ethnography, and skills of social scientists are becoming ever more relevant in the research and development teams. Think of autonomous vehicles (projects by Nissan and Volvo Cars already involve anthropologists), supporting energy efficiency in buildings (Horizon 2020 project MOBISTYLE, where anthropologists contribute to the development of ICT services), or our PEOPLE project, where teams of students in anthropology, sociology, and related fields work with industry and academic mentors on four different case studies in energy efficiency and sustainable living. Not to mention the vast amount of work done by social scientists and researchers in the field of people-centred (or human-centred) design. Ethnography seems to be the thing – outside academia and in interdisciplinary research and development teams.

Ethnography 2.0 – Going Interdisciplinary and Digital

But how do we do ethnography in these non-traditional settings? Does it get altered when we are dealing with digital stuff and large data sets? Can it still be called ethnography, when there is only a fraction of the traditional 18 months available for fieldwork, for example when anthropologists are only one small part of a large interdisciplinary R&D project? How do anthropologists and other social scientists fit into the development teams alongside engineers? These and other questions were addressed at the symposium Ethnography 2.0, which brought together researchers from a number of disciplines and research fields, addressing data mining, artificial intelligence, knowledge technologies, and bioinformatics, among other things.

Dan Podjed, interdisciplinary research teams, are you serious, anthropology, method, ethnography
Dan Podjed (ZRC SAZU): Are You Serious? Initial reaction of engineers to an explanation of how anthropologists will contribute to an interdisciplinary R&D project.

Two members of the Slovenian PEOPLE team were presenting at the conference. In his paper, Dan Podjed (ZRC SAZU) presented the PEOPLE project and other interdisciplinary projects, in which he has been working as an anthropologist among natural scientists and engineers. He explained how the research team’s initial reaction to his presentation of the anthropological approach was: “Are you serious?” (The usual suspects: too small sample, statistical insignificance, impossible to provide useful development guidelines.) However, as he explained, the conclusions, derived from the ethnographic research, were eventually taken very seriously by the engineers and successfully incorporated in the solutions developed in the projects. In interdisciplinary R&D projects, as well as in industry or business settings, anthropologists need to be versed in explaining the value of ethnography – repeatedly and clearly. Dan discussed how anthropological methods are often transformed and adapted in interdisciplinary contexts.

Ajda Pretnar is participating in PEOPLE as a PhD student in anthropology, and she is also a researcher at the Faculty of Computer and Information Science, Bioinformatics Laboratory. She is researching machine learning, computing analysis in anthropological research and the use of new research methods in social sciences. In her paper, she presented the anthropological approaches to digital data and defined concepts such as data ethnography, digital anthropology and computational anthropology. She presented the possibilities of using machine learning and data mining in anthropology, drawing on her research on uncovering behavioural pattering from sensor data in a smart building (see also Slovenian case study) and classification of hayracks. Ajda also led a workshop on Data Ethnography and presented some practical examples for using new computer-based methods of data analysis in anthropology.

Etnografija 2.0, ethnography, data ethnography, digital anthropology, symposium, audience, interdisciplinary
The interdisciplinary audience.

Rajko Muršič (Faculty of Arts, University of Ljubljana), Blaž Bajič and Sandi Abram (University of Eastern Finland) talked about sensory ethnography and presented the project SensotraAljaž Košmerlj (Jožef Stefan Institute) talked about artificial intelligence and presented the current projects undertaken at the Laboratory for Artificial Intelligence. Darja Fišer (UL and Jožef Stefan Institute) discussed the management of research data, in particular language data in ethnographic research, following the principles of open data and FAIR guidelines (Findable, Accessible, Interoperable, Reusable). Jana Šimenc (ZRC SAZU) talked about digital medicine, the relation between digital technologies and health, and presented her ethnographic research of digital health in Slovenia. Marko Robnik-Šikonja (Faculty of Computer and Information Science, UL) presented the idea of deep learning and its use in analysis of image, sound and textual data, as well as its limits and the related ethical questions. As an example of analytical methods using machine learning he presented the (rather controversial) case of inferring a person’s sexual orientation by processing their facial images (read about the topic in this Guardian article).

The Future is Already Here

The mix of presenters, including anthropologists, linguists, computer and information scientists, proved to be a potent combination. The discussion turned around the issues of how qualitative and quantitative methods are complementary and bring valuable insights when combined. Even more so, when the research teams are dealing with digital data, our technological present and potential futures, as demonstrated by a number of examples from interdisciplinary research projects.

Inevitably, the anthropologists in the room raised the question of ethics in dealing with sensor data or artificial intelligence, for instance. One of the commenters explained that while she initially reacted with caution and worry to the idea of anthropologists using machine learning or data mining as tools in their research, she now felt that it was important that social scientists are involved in the development of AI and new digital technologies from the very beginning. After all, anthropologists will be those asking the uncomfortable questions concerning ethics and the wellbeing of people. Apparently, we should ask those questions sooner rather than later.

Ethnography 2.0, symposium, Dan Podjed, Ajda Pretnar, data ethnography, digital anthropology, inerdisciplinary, data mining
Proof that anthropologists can handle technologies (and support them with books). Featuring: Ajda Pretnar and Dan Podjed.