Blogpost: Sofia Caldeira: When working on 'the digital' is something other than Digital Humanities

As part of our recent Digitial Humanities Doctoral School's programme, participants were asked to write a blogpost capturing their experiences with the digital humanities. In the first blogpost of the series, we welcome Sofia Caldeira, a PhD-candidate at the Department of Communication Sciences, Ghent University. Sofia is conducting a research project on representations of femininity on Instagram and women’s glossy fashion magazines, funded by the Fundação para a Ciência e Tecnologia (FCT).


When working on 'the digital' is something other than Digital Humanities: Exploring the potentialities and limitations of the field in the context of my PhD project “From Women’s Magazines to Instagram: A qualitative analysis of photographic representations of femininity in the age of self-representation”

I feel that I should start this text with a personal disclaimer. Despite the very digital nature of my topic – the photo-app Instagram – and of my participation on this Digital Humanities blog, I am not a digital humanities scholar. In fact, I wouldn't even call my self a humanities academic, in any clear-cut, classical sense. My somewhat erratic academic path – one that started in Fine Arts, passed through Cinema, Visual Anthropology, and now Communication Studies – has left me hesitant to claim any label. I am merely an outsider, curious to see what the then-unknown field of the Digital Humanities had to offer.

My approach is one that does not easily fit with the ones usually associated with Digital Humanities – which brings to mind digitized collections, digital tools and automated methods of analysis, data visualization, etc. When considering visual social media, like Flickr or Instagram, Lev Manovich (2012) saw it as ripe with possibilities for digital study. These platforms seem to invite a big data approach, with their billion of photographs uploaded daily, countless captions, comments, tags, and other metadata information waiting to be collected. Yet, my research on this same object is not focused on software or code, nor on the social network aspect of Instagram; instead it focus on small qualitative-oriented samples, not using any large scale automated data-mining tools. My approach is still very much rooted in the traditional ways, focusing instead on what Manovich (2012) called the “deep-data”, a comprehensive study of a few cases.

The same research object and interests – Instagram and photographic self-representation – has, of course, been approached in ways much more easily associated with the Digital Humanities tradition. An example of this is the Selfie City project (Manovich et al. 2014), a large scale, big data approach to the the selfie-production of five different cities worldwide, that sought to uncover cultural patterns in large sets by employing automated visual analysis (albeit, complemented by human judgement, as the algorithms used are still liable to mistakes).

Projects such as this take advantage of the digital tools of data collection and analysis to cover previously unimaginable amounts of information. Using the API's made available by the social media platforms, or using methods of PHP data collecting, one can easily and quickly collect large amounts of data. And yet, as Manovich (2012) himself noted, these tools are not without its limitations. The apps can limit the type and amount of data researchers can have access to, and using these tools still implies a certain level of “tech-savviness” (or at least having a tech-savvy friend willing to help). 

A clear example of these limitations was the decision of Instagram to change its policy for API use in July 2016. It limited the access to data only to those seeking to develop new apps (and particularly, apps that will ultimately be economically beneficial to Instagram), thus effectively locking most academic researchers out. And as the use of data scrapping tools is usually frown upon by app owners, researchers are left in a somewhat weird position when it comes to automated data collection from Instagram.

Furthermore, the idea of collecting data from social media is a fertile ground for ethical dilemmas. Some scholars, like Lev Manovich (2012) minimize these ethical concerns by trusting the good sense of academic researchers and defending that we don't need to ask research subjects permission to study and collect their social media production “(...) since they themselves encourage us to do so by making all of this data public.” Yet, I believe that the highly personal and identifiable nature of these materials calls for a more careful consideration and approach. Although people often leave their social media profiles public (and as such accessible to all), we must keep in mind that “public” is the default setting of Instagram, and that many people do not even consider this or bother to change it after setting up their accounts. This does not mean, however, that they do not have a certain expectation of privacy when using social media, seeing it as a way to interact with friends and acquaintances, and not with a wider potential audience. There are no universal guidelines. These ethical considerations depend not only of personal sensibilities, but also from the kind of research one intends to do. In my case – after trying to imagine myself in the position of the research subject, being studied without my explicit knowledge or permission – I've opted to request the informed consent of each participant before collecting any data, making sure they understand the research and its aims. 

While these tools for automated data collection, whenever available and authorized by the research participants, will undoubtedly bring advantages to my research process, making it much less time consuming, I still remain on the fence regarding the idea of automated data analysis. Manovich (2012) himself acknowledged that computers can't do the same kind of nuanced interpretation and are not supposed to “replace” human experts. An over-reliance on automated analysis lends itself to the critiques of loss of “humanist sensibility”, “technological anonymity” and even of mere “number-crunching” that are occasionally made to the field of the Digital Humanities (Eyers 2013).

My approach couldn't be more diametrically opposed from this idea of number-crunching. I've chosen to approach Instagram in a eminently qualitative manner, using small samples that have no claims of being representative or generalizable, but that are rather illustrative. The methodologies I am employing are also those of the social sciences, such as a 'textual' analysis of the photographs shared on Instagram, employing close reading strategies, and virtual ethnography, transposing the practices of participant observation to the digital field and seeking a naturalistic approach, centred on the subject and on the “lived online experience” (Kozinets, Dolbec & Earley 2014).

Even my focus on questions of gender, identity and politics of representation seems to, unwittingly, go against some of the general criticism that was pointed at the Digital Humanities of having a “lack of attention to issues of race, class, gender and sexuality” and a “absence of political commitment” (Gold 2012). Working from the tradition and the theoretical framework of feminist media studies, my work arises as inevitably engaged.

In the end, I've come to realize that simply using a computer and looking at Instagram is not enough to make you a Digital Humanist. Although this field has shown me several possibilities (some that I had never previously considered) and presented me with tools that have the potential to make my data collection process much less time consuming, my research remains firmly grounded in a different academic tradition. I've made my peace with the fact that I fall in that category that Kathleen Fitzpatrick (2012) defined as scholars who work with digital materials, but who remain outside the tradition and assumptions of Digital Humanities.


Eyers, T. (2013) 'The Perils of the “Digital Humanities”: New Positivisms and the Fate of Literary Theory'. Postmodern Culture: Journal of Interdisciplinary Thought on Contemporary Cultures, 23 (2). 
Fitzpatrick, K. (2012) ’The Humanities, Done Digitally'. In: Gold, M. (ed.) Debates in Digital Humanities.
Gold, M. (2012) ’The Digital Humanities Moment'. In: Gold, M. (ed.) Debates in Digital Humanities.
Kozinets, R., Dolbec, P.  & Earley, A. (2014) “Netnographic Analysis: Understanding Culture through Social Media Data" In: Flick, U. (ed.) The SAGE Handbook of Qualitative Data Analysis. London: SAGE Publications.
Manovich, L. (2012) 'Trending: The Promises and the Challenges of Big Social Data'. In: Gold, M. (ed.) Debates in Digital Humanities.
Manovich, L. et al. (2014) Selfie City.