Justus Uitermark is an Associate Professor of Sociology at the University of Amsterdam. In our interview he reflected on some of his most recent work, which looks into the geographical dimension of social media activity. In particular, he discussed the paper “Reassembling the city through Instagram”, written with Dr. John Boy from Leiden University. This paper looks into the ways that people represent the city in social media and how those representations feed back into people’s experiences of the city. Uitermark remarked, digital methods “allowed a way into worlds that were quite alien to us”. The big corpus of data, which in the study discussed was comprised of over 400.000 geo-tagged Instagram posts from Amsterdam, enabled the researchers to comprehensively map different groups of Instagram users, which “really wouldn't have been possible if we had relied on different methods”. Thus, he believes that what digital methods affords researchers is to “get really comprehensive’’ and to consider the entire population of users while also providing them with a way to select sites and groups for further studies.
Uitermark highlighted three main contributions that the use of digital methods, in combination with other approaches, can provide to the study of culture and society. First, he believes that social media research is a step forward to a more relational research, as envisioned by influential sociologists like Pierre Bourdieu and Norbert Elias. These authors, he explained “ask us to view the world relationally. To not look at individuals as autonomous beings but to consider them as part of larger sets of relationships”. While he cautions that social media data does not allow scholars to grasp the type of relationships that Bourdieu and Elias were interested in, it does make their ontology applicable and enables researchers to “see who is connected to who, to understand individual behavior within a larger relational context”.
Secondly, Uitermark expressed that he is very positive about the ability of digital methods to grapple with representations and to do it at larger scale. This allows for what he calls “serendipitous discovery”, that is, being able to “see things that you didn't anticipate”. While the use of social media data can be perceived as excessively positivistic, professor Uitermark believes that this type of data is also suitable for more exploratory research that makes visible new sites and groups of study and their relationships.
In particular, the inductive approach to social media data encouraged by Dr. Uitermark is exceptionally relevant for globalization research. In the case study discussed above, for instance, the ability to map relationships and to zoom in and out exposed how “many of the interactions are not within the city but beyond its boundaries” even when the focus of their work was on Amsterdam. This, he proposes, raises really interesting questions on how to delimit an object for analysis in an interconnected world where, partly because of social media, place experience is not bound to a particular locale. Hence, in their study of the city, social media data compelled them to have a more complex notion of place, which requires scholars “to think in terms of the global or at least trans-local circulation of images and communication”.
Finally, Professor Uitermark explains that social media data provides scholars with a tool to provisionally or partially test theories. Moreover, having a more direct access to the studied connections and networks, he points out, makes it “even possible for us to understand phenomenon that we are not intimately familiar with” which he believes “is a huge advantage, it is something that we should use with caution, but also with considerable enthusiasm and optimism.”
Using digital methods, however, is not without limitations, as Dr. Uitermark points out in our conversation. On the one hand, for the work discussed, data collection that relies on posts with geo-coordinates leaves out many of the images circulating on Instagram, since only a small proportion of posts are geo tagged. Hence, Dr. Uitermark remarks, a lot is lost and, to add fuel to the fire, researchers cannot know what posts they missed, “so we don't know how our sample was biased”.
On the other hand, and maybe more importantly, even if researchers had access to complete data, Dr. Uitermark points out that“still you would end up making inferences about the meaning of the data”, which he believes is a “huge limitation”. To understand social media data and appreciate its meaning and significance, he explains, researchers must understand it as part of practices, which “never come into view when you look at platforms alone”. This makes this type of research prone to biases, since “people have all sorts of assumptions about these practices and this is something highly problematic”. This issue would be solved, Dr. Uitermark suggests, by studying those practices themselves, “but this raises the question of research capacity and research competence because it is a lot of work to do and it requires a range of different competences”.
Another issue that arises when using digital methods is that “having data at your fingertips can result in a sort of very ephemeral, very shallow, while also very utilitarian relationship to the object of study”. This is particularly problematic, professor Uitermark suggests, when studying globalization. Using social media data, he explains, “can be particularly deceptive because it might give you the impression that you can grasp everything, because you have all this data from around the world”. In this scenario, a mixed methods approach and a sensibility for local nuances become essential: “If you start looking closer in particular localities, you would become more aware of the limitations”.