The monads of Facebook: A comparative study of two digital methods approaches

Olesen, Asger Gehrt

4. term

Techno-Anthropology, Master

2018

2018-06-01

70

Vores sociale liv foregår i stigende grad både offline og online. Når vi interagerer digitalt, efterlader vi spor, der gør det muligt at undersøge sociale relationer på nye måder. Dette speciale sammenligner to metodiske tilgange til at studere det sociale på digitale platforme og viser, hvordan de adskiller sig empirisk. Den ene er den emneorienterede tilgang, som bruger kontroverser til at se, hvordan aktører danner alliancer ved at engagere sig i de samme problemstillinger. Den anden er den medieorienterede (medium-centric) tilgang, som udnytter platformenes indbyggede måder at organisere interaktioner på for at forstå, hvordan netværk opstår. Begge tilgange hører under digitale metoder, et felt med rødder i aktør-netværks-teori (inden for videnskabs- og teknologistudier) og i mediestudier. Tidligere forskning blander ofte tilgangen, hvilket gør det svært at se, hvad hver metode bidrager med, hvor de ligner hinanden, og hvilke empiriske konsekvenser valget af metode har for repræsentationen af sociale relationer. Med udgangspunkt i Gabriel Tardes begreb monader undersøger specialet denne problematik. Monader forstås her som enheder, der kan forbindes—ikke kun mennesker, men alt der kan knyttes relationer mellem. Specialet argumenterer for, at Facebook kan betragtes som et netværk af monader: Venner indgår i hinandens monade, og brugere bliver også del af monader, når de forbinder sig til grupper og sider gennem kommentarer, likes og delinger. Problemformuleringen lyder derfor: Hvilke empiriske forskelle kan observeres i, hvordan den emneorienterede og den medieorienterede tilgang registrerer forbindelser i monadiske lister? Og hvilke implikationer har det for social teori? Spørgsmålene besvares i et komparativt studie af en konkret Facebook-sag: uenigheden om, hvorvidt HPV-vaccinen forårsager bivirkninger. Analysen bygger på to datakilder: et nyt datasæt med opslag fra 72.000 offentligt tilgængelige danske Facebook-sider og et datasæt med manuelt udvalgte opslag relateret til HPV-kontroversen. Begge datasæt er indsamlet via Facebooks API. Specialet bidrager på tre måder. Metodisk opstiller det for første gang en klar operationalisering af både den emneorienterede og den medieorienterede tilgang, som muliggør en gennemsigtig sammenligning. På den baggrund konstrueres to sammenlignelige netværk, der repræsenterer de to tilgange. Empirisk viser specialet store forskelle i, hvordan tilgange repræsenterer det sociale i HPV-kontroversen: Overlappet af monader på tværs af de to netværk er lavt; de tre grupperinger, der identificeres i hvert netværk, ligner i begrænset grad hinanden; og de aktører, der fremstår centrale i det ene netværk, er ofte perifere i det andet. Det hænger sammen med, hvordan centralitet identificeres: Den emneorienterede tilgang belønner monader, der bruger det relevante vokabularium, mens den medieorienterede belønner monader, der samler mange interaktioner fra aktører, som også interagerer andre steder i netværket. Teoretisk betyder det, at man ikke kan antage, at de to tilgange giver samme fremstilling af en kontrovers. Valget af metode er samtidig et teoretisk valg om, hvad der tæller som en social forbindelse.

Our social lives increasingly unfold both offline and online. As we interact digitally, we leave traces that make it possible to study social relations in new ways. This thesis compares two methodological approaches to studying the social on digital platforms and shows how they differ empirically. The first is the issue-oriented approach, which uses controversies to see how actors form alliances by engaging with the same problems. The second is the medium-centric approach, which leverages platforms’ built-in ways of organizing interactions to understand how networks emerge. Both belong to digital methods, a field rooted in actor-network theory (within science and technology studies) and in media studies. Prior research often mixes these approaches, making it hard to see each method’s distinct contribution, where they overlap, and what empirical consequences the choice of method has for representing social relations. Using Gabriel Tarde’s concept of monads, the thesis examines this challenge. Here, monads are understood as units that can be connected—not only humans but anything that can be linked. The thesis argues that Facebook can be seen as a network of monads: friends are part of each other’s monad, and users also join monads when they connect to groups and pages through comments, likes, and shares. The research question is therefore: What empirical differences can be observed in how the issue-oriented and medium-centric approaches record connections in monadic lists? And what are the implications for social theory? These questions are addressed through a comparative study of a concrete Facebook case: the disagreement over whether the HPV vaccine causes side effects. The analysis draws on two data sources: a new dataset of posts from 72,000 publicly available Danish Facebook pages and a dataset of manually selected posts related to the HPV controversy. Both datasets were collected via Facebook’s API. The thesis makes three contributions. Methodologically, it offers, for the first time, a clear operationalization of both the issue-oriented and the medium-centric approach, enabling transparent comparison. Based on this, two comparable networks are constructed to represent the approaches. Empirically, the thesis shows substantial differences in how the approaches represent the social in the HPV controversy: there is low overlap of monads across the two networks; the three groupings identified in each network are only weakly similar; and actors that appear central in one network are often peripheral in the other. This stems from how centrality is identified: the issue-oriented approach rewards monads that use the relevant vocabulary, while the medium-centric approach rewards monads that accumulate many interactions from actors who also interact elsewhere in the network. Theoretically, this means one cannot assume the two approaches will yield the same account of a controversy. Choosing a method is also a theoretical choice about what counts as a social connection.

[This abstract was generated with the help of AI]

Digital methods ; Tarde ; ANT ; Monads ; HPV

Download
View record in AAU Student Projects