Selfiecity is a research project led by Lev Manovich which attempts to make sense of the multitude of selfies posted to Instagram. Over the course of a week between 5th December to 11th December 2013, photographs shared on Instagram from 5 cities across the world (New York, Bangkok, Moscow, Sao Paolo and Berlin) were selected before being sorted into images that showed a “true selfie” – that is, a photograph of a single person taken by themselves. From these a selection of 1000 images was chosen from each city and then classified using face-analysis software to measure facial characteristics such as size, orientation and mood as well as gender and age. (A detailed description of the process can be found in this essay by Alise Tifentale and Lev Manovich here.) Dominikus Baur, a contributor to the project, makes this analysis of why the selfie is worthy of such meticulous study:
“selfies are a fascinating target for analysis: especially in their Instagram incarnations, they’re a massive data set offering a unique glimpse into the public psyche. Selfiecity is our stab at what studies of such data can look like, methodology – and result – wise.” (Baur, 2014)
The Following outputs were made with the data and images:
Details of the demographics of people taking selfies, their poses and expressions. The first finding is perhaps the most surprising – only 3-5% of the images from the initial selection were “true selfies”. (Given that this study was completed in 2013 however, I wonder if this would still be the case today.) Other assumptions were proved to be true however, such as significantly more women take selfies and the majority are taken by younger people with the median age being 24, although more older men post selfies than women. (See this essay by Mehrdad Yazdani for further details.)
Blended Video Montages:
Video montages made from images from each of the 5 cities can be found on the Selfiecity website here. Each video presents 640 selfies from each city with the images aligned at eye level and the new ones fading on top of the old giving an abstract representation of both the individuals and their context. Moritz Stefaner (2014) describes the effect as “a dynamic, morphing ‘aggregate city face.'” Tifentale and Manovich explain further:
“This visual strategy is designed to create a tension between individual shots and high-level patterns. We don’t show each face by itself. But we also don’t superimpose all faces together – which would create a generic template. Instead, we show something else: a pattern and individual details at the same time.” (Tifentale and Manovich, 2014)
These, histogram-type visualisations of the individual images, seek to show the pattern and distribution of genders, ages and expressions in each of the cities observed. Shown as graphs composed of individual images, they allow the viewer to explore the interplay between similarities and differences within each set.
This is an interactive visualisation app that allows site visitors to explore the data collected in the project and filter the photos by city, gender, age and a number of face measurements such as pose and head tilt. The functionality of this part of the project is extremely intuitive and I found myself playing with the settings for some time and exploring the varying outputs of individual images.
Selfiecity is a fascinating project both in terms of research rigour and artistic output. Although the project seeks to provide empirical data to either prove or refute commonly held notions about the the selfie, it can only be a snapshot of a particular moment in time and I wonder what differences would be found if the project was rerun today. The difficulty in being able to categorise the selfies is discussed in many of the essays on the website and is another interesting point to consider. Deciding whether an image was a true selfie was often contentious and prone to disagreement, eventually relying on a system of voting. The essays on the development of the selfie and their social and cultural relevance available on the website are also fascinating insights into a complex and far reaching subject that has fired my imagination to research further.
- Lev Manovich website
- Selfiecity website
- Phototrails website (an earlier Instagram project by Manovich)
- Alise Tifentale website
Baur, D. (2014) Data Visualisation: Selfiecity. Available at: http://do.minik.us/blog/selfiecity (accessed 6th July 2020)
Hochman, N. (2014) Imagined data communities. Available at: http://d25rsf93iwlmgu.cloudfront.net/downloads/Nadav_Hochman_selfiecity.pdf (accessed 5th July 2020)
Losh, E. (s.d.) Beyond biometrics: feminist media theory looks at Selfiecity. Available at: http://d25rsf93iwlmgu.cloudfront.net/downloads/Liz_Losh_BeyondBiometrics.pdf (accessed 5th July 2020)
Stefaner, M. (2014) The design of Selfiecity. Well-Formed Data. Available at: http://well-formed-data.net/archives/996/selfiecity (accessed 6th July 2020)
Tifentale, A. (2014) Making sense of the “masturbation of self-image” and the “virtual mini-me”. Available at: http://d25rsf93iwlmgu.cloudfront.net/downloads/Tifentale_Alise_Selfiecity.pdf (accessed 5th July 2020)
Tifentale, A. and Manovich, L. (2014) Selfiecity: exploring photography and self-fashioning in social media. Available at: http://manovich.net/index.php/projects/selfiecity-exploring (accessed 23rd May 2020)
Tifentale, A. and Manovich, L. (2016) Competitive photography and the presentation of the self. Available at: http://manovich.net/index.php/projects/competitive-photography-and-the-presentation-of-the-self (accessed 23rd May 2020)
Yazdani, M. (2014) Gender, age, and ambiguity of selfies on Instagram. Software Studies Initiative Blog. Available at: http://lab.softwarestudies.com/2014/02/gender-age-and-ambiguity-of-selfies-on.html (accessed 6th July 2020)