The Geography and Gaze of the Selfie.

Although self-portraits have an important place in the history of artistic expression and have been in vogue for centuries, the selfie – the digital portrait of oneself, most often taken by holding a camera phone at arm’s length and subsequently shared through social media – has captured the global public imagination with great force over the last decade. Early selfies were often taken with webcams hooked up to desktop computers and are more stationary in character. They are taken within the home, within the intimacy of the living room or bedroom, and mostly focus exclusively on the photographer’s face. But the prevalence of camera- and internet-enabled smartphones has brought the selfie into the outside world. A selfie today is as much about the photographer’s surroundings as it is about the photographer themselves. Considering how selfies are intricately linked with various social media platforms, the selfie cannot be considered to be just a visual object or cultural representation. Instead, of just being a cultural product, the selfie is intricately tied up in producing social media culture . Selfies are viewed, ‘liked’, commented on and shared onwards by followers and friends on social media platforms as well as by more traditional media. Since smartphones have enabled the selfie to untether itself from the living room webcam, naturally the geography of the selfie and how it produces representations of places becomes an important and intriguing subject. This amplified further by the fact that many selfies are explicitly geotagged. That is, users of social media platforms such as Instagram and Twitter choose to add a specific geographic location to the selfie they are uploading. With this implicit indication that the location of that photo is somehow important, we ask ‘what is the geography of the selfie?’.To answer these questions, this article uses a global dataset of geotagged selfies posted to Twitter since the summer of 2012 (cf. Poorthuis et al., 2016). By matching based on keywords and hashtags related to selfies, a dataset of ~10 million tweets and their corresponding images is extracted. The empirical section on the article consists of two parts. In the first part, characteristics of these selfies are compared with random samples that represent all tweets and all tweets with images attached, respectively. These datasets are mapped and visualized to analyze the geographic distribution of selfie production.