Making the Invisible Visible

Context

Before starting this project, we were asked to pick a topic from a list. We chose online censorship, specifically of images, as it presented an interesting opportunity for qualitative research.

Online censorship involves controlling or removing specific public content by governments or platforms aiming to regulate it. As visual content production grows exponentially, these policies have become increasingly detailed and numerous. This phenomenon impacts more users, raising significant questions: Why is some content censored? Who is responsible for the censorship? Where is the boundary between appropriate and inappropriate content? Censorship can sometimes act silently, leaving users unaware, while at other times it is blatantly evident, prompting users to seek ways to respond or bypass it.

We were tasked with using digital methods to conduct the research, and we aimed to come up with innovative ideas on how to study images with them.

Process

1. Defined areas of interest and established identity guidelines for visual consistency.
2. Designed research questions and protocols, and collected and analyzed data for each.
3. Designed tailor-made and non-conventional visualizations for each research question, ensuring the inclusion of collected images.
4. Selected the main findings for each protocol and developed a website to present the results.

My team and I developed eight different research questions to explore the theme of image censorship online: (1) What is Google’s ranking of censored places on Google Maps and what kind of places are they?, (2) How did the censorship of the researched places online change over time and which are the techniques used to censor them on Google Earth Pro?, (3) Which are the censored artists and artworks on social media that are more mentioned on Google Search?, (4) Does online censorship on social media treat women’s and men’s bodies equally?, (5) Which are the techniques of self-censorship used by erotic magazines to deal with Instagram policy?, (6) Which are the sub-topics related to freethenipplemovement on Instagram?, (7) Which are the features mentioned in the content moderator’s job offers, how should candidates be and what are they expected to do?, and (8) How is the distribution of the content moderator’s job offers around the world, and which are the required languages per country?

These questions delve into three levels of depth regarding censorship:

1. Types of censored content and the visual languages employed.
2. Responses elicited by censorship, including anti-censorship movements and methods users employ to circumvent it.
3. The individuals behind the phenomenon.

While I advised on collecting and analyzing data for questions 4 and 6, my focus was on questions 3 and 5, which I managed entirely.

I began by collecting censored artwork and artist names from the first 100 relevant Google results. This provided a dataset on which artworks and/or artists were censored on which platforms and how. I then created an image database of the artists’ works or specific artworks from Wikimedia or Google Images.

I designed a treemap visualization to make it clear which artworks were most frequently mentioned, creating a visual censorship gallery where each artwork is framed by the type of censorship it faced.

Noticing how some independent magazines self-censored their content to share on Instagram, I thought it would be interesting to map these techniques. I manually collected the names of erotic magazines via a Google search and further filtered them to focus on those actively using Instagram.

Using an Instagram scraper tool, I collected the first 200 images from each account, stored them in a database, and analyzed them to catalog the types of censorship applied.

I designed the visualization to resemble an archive, where each self-censored picture is represented by a small rectangle, grouped by magazine, and colored according to the censorship technique. Hovering over a rectangle reveals the image, mimicking the physical navigation of a register. This visualization aimed to immediately show which magazines had the most self-censored images and the most used techniques.

Outcome

“Making the Invisible Visible” is a website that presents our research on online censorship of images. This digital report includes eight research protocols and related visualizations, along with explanatory texts, main findings, and protocol schemes.

We aimed to provide a comprehensive view of this phenomenon and provoke reflection on our findings.