Business Fish goes all in: affect-images and the Facebook machine

8 May , 2015  | by:

In this brief piece, I’m going to address the culture of image responses in online communication, and look at how and why I think Facebook has incorporated its own system. Within the last two years Facebook has provided a new feature in its chat and thread systems: the ability to comment or post using small images called ‘stickers’. Finding a precise date on the inclusion of stickers into the Facebook social media ecology is difficult because the inclusion hasn’t been marked by much in the way of press releases or other trumpeting. Stickers subtly became a part of the everyday use of Facebook without much fanfare, neither changing extant services nor replacing existing ones. We collectively woke up one day, logged-in, maybe noted a new button in our chat windows and then perhaps thought in passing “was that smiley face there yesterday?” Perhaps we used them, perhaps we did not, but they were now here to stay.

Earlier in 2015, Facebook stickers seemed to obtain their own cultural niche. The well-known and well-loved sticker of professionalism, Business Fish, was the subject of articles in The Guardian, Junkee, and Rhizome, and a parody Twitter account began operation on February 7th this year. These thinkpieces are less of a critical reflection on stickers tout court, and more of an empathetic response to what Business Fish *means*. That, somehow Business Fish has tapped into a cultural niche that exists nowhere else. I think that perhaps, yes, Business Fish does represent something significant in the cultural zeitgeist of depressed, piscine, white collar office workers, but to borrow from Canadian media studies, this matters less than the role that stickers have over all. The issue is that Business Fish is only one of a large number of stickers available over Facebook’s system, and I want to raise up the possibility that these stickers need to be considered in a larger scope.

Image responses are a large part of internet culture for years. From 4Chan to Jezebel, 9Gag to MySpace, the use of images as responses is a significant part of how people practice communication online communication. Emoticons are a variation of this, and have themselves existed for many decades. Using images as representing specific, select responses to certain emotional states is a far more recent development, due probably to increasing global internet throughputs and greater access to broadband connections globally. Perhaps, oddly, we could think of this tendency as a part of a maturing of a digital communications culture, despite the immaturity of many of the images. I would pose that these images are affect-images, in the Deleuzean sense (among other ways we might analyse them) – these are images deliberately chosen for how they pass on an affect: i.e. images that pass on a sense of an involuntary tic or gesture that comes unbidden to the viewer, often in direct response to a particular comment or idea. As Deleuzean affect images they are facified – while not literally a face, they possess some of the qualities of a face: a gloating look, the raised eyebrow of success, glee and laughter, and so on. These affects are different from the emotional language that we use in everyday life – the image of a raised eyebrow implies a different way of responding to a situation than comments such as ‘wow’ or ‘OMG’. They’re not exclusive, but they do express different aspects of subjective experience.

Response images are numerous. Image boards have been around for a good decade or so, and these relied on users uploading images from their own computers. The uploaded image would then be assigned a random number, which would then invariably be downloaded by some other enterprising image board community member. This circuit, the image-post-image’ circuit, would lead to one image being redistributed numerous times with a different file name, or simply renamed by each user. This means that the same image file is hard to track, without doing some sort of time-intensive reverse-image-lookup or setting up equally time intensive MD5 hashes. When this image-response practice began to take place in Facebook, this meant that a not insignificant proportion of user communication was being conducted in terms of images that Facebook could not efficiently track. Cue stickers.

Facebook’s sticker images seem to sit somewhere between a meme and a response gif, with strong tendencies towards those horrendous, cacophonous, e-cards of the mid/late-90s (“page under construction”, etc). Each image expresses a singular situation or emotion. Many are animated, and most convey a specific emotion or response. Many of the sticker images are somewhat unsatisfying when used. ‘Oakley In Action’, for instance, is a bright green owl which has some of the aesthetic sensibility of Microsoft Clip art (although perhaps not to the degree of the ‘Dance Party’ set), and seems to revel in the denotative layer of meaning. Like the worst political cartoons, Oakley seems to struggle to reach any broader connotative meaning, and really strips back the possible interpretative systems available to the reader. Oakley’s role is not to produce meaning, but drives us to limit what we’re talking about to highly specific uses, and they’re probably ironic. Oakley on a skateboard. Oakley yelling through a megaphone. Oakley listening to music. These are all amazingly non-evocative images. They don’t attempt to connect to any greater meaning than skating, yelling, and listening, and their use is highly limited.

So, finally, my research question: why does Facebook provide those emoticon sticker packs when the system readily accommodates for image files to be uploaded or linked to in its service? People could already respond to posts, threads, and messages by adding images and links to gifs on Facebook at the time that Stickers were released.

If this question is directed at people, then the first answer is usually economic: ‘because they can sell sticker packs’ (now irrelevant because the stickers are free) and the second is usually pragmatic: ‘they’re just improving the service’. Certainly, finding a way of facilitating mobile users to engage in image responses in the same way as computer-based users is probably important. It’s a lot easier to surf the net on a computer to find the right response gif or a particularly nuanced version of ‘know that feels guy’. But neither answer is very satisfying. Here is my answer: sticker packs are a deliberate attempt to replace a part of internet culture with Facebook’s own equivalent system which is easier to use and standardized, and thus easier to track.

Stickers are standardized in terms of their metadata profile. There is no documentation for stickers on Facebook’s various APIs, but inspecting any sticker via browser-based inspection tools reveals that each sticker has a 15-digit reference number, by which the sticker can be easily tracked. This simple 15 digit number means that Facebook holds a standardization of images for its stickers, which means that the stickers can be tracked, meaning that affective profiles can be built from user’s comments and responses on threads and messages.

If you look at the methodological schema that Facebook uses for its text analysis of user content you can see that there are 464 words that have been assigned an ‘emotional’ status. Facebook is now expanding that list to include a large number of images which allow them to get a more granular analysis of the emotional content of responses. Furthermore, I would argue, baselessly, that the stickers contain a degree of affective communication. Most of the stickers are not expressive of specific emotional content, and are rather images of bodily reactions which can be distributed to many different situations in place of purely emotional descriptions. In this sense, stickers are not a replacement for other types of emotive communication, but a means by which Facebook can expand the data-capture net wider, and grab aspects of people that have previously been difficult to grasp.

As indicated by their largely maligned study, “Experimental evidence of massive-scale emotional contagion through social networks”, the study of emotional impact and expression over social media is an important research area for Facebook’s research teams. While I have no specific proof, I believe it would be trivially easy for Facebook to integrate the use of stickers into its data-tracking programs in order to build a finer degree analysis of users in terms of aspects that are not normally measurable.

, , ,

Comments are closed.