Digital Traces of the Persona through Ten Years of Facebook
Why this work is in the frame
A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.
Bibliographic record
Abstract
When I think, rarely, about the articulation of the set of traces that I am leaving, I have the immediate apprehension that it is not the real me that’s out there on the Web. I know the times when I have censored myself (oh problematic concept!) and when I have performed actions to complement—and frequently to confound—a trace. […] Taken globally, the set of traces that we leave in the world does without doubt add up to something. It is through operations on sets of traces that I understand an event that I take part in. (Bowker 23) Over the past decade, Facebook has become integrated into the everyday lives of many of its 1.28 billion active users to the point that Facebook can no longer be considered “new media.” The site is driven by the “disclosures” (Stutzman, Gross and Acquisti) users make on the site—by uploading photos, writing status updates, commenting on posts made by others, sharing news items, entering biographical details, and so on. These digital traces of life are archived by default, persisting indefinitely as etches in Facebook’s servers around the world. Especially for young users who have grown up using Facebook, significant parts of their social and cultural lives have been played out on the site. As spaces in which the persona is enacted and made visible, social network sites like Facebook also effectively capture growing up stories through a chronicle of mediated, transitional experiences: birthdays, graduations, the beginning (and end) of relationships, first jobs, travel, and so on. For these reasons, Facebook also comes to serve as a site of memorialisation for users who have passed away. To mark its tenth anniversary (2014), Facebook drew attention to the great depth and wealth of experiences users had traced upon its pages through the release of one-minute “look back“ videos, chronicling the life of individual users over their time on Facebook. These videos have become short manifestations of the personas presented on the site, crafted through an algorithmic selection of critical moments in the user’s life (as shared on the site) to tell that user’s story. To turn Bowker’s musings in the above quote into a question, what do these sets of traces that we leave in the world add up to? In this article, I undertake a critical reading of Facebook’s look back videos to argue that they serve as the strongest reminder yet about the function of Facebook as memory archive. I draw on several sources: my own analysis of the structure of the videos themselves, the Facebook corporate blog describing the roll out of the videos, and the public campaign played out on YouTube by John Berlin to have a look back video generated for his deceased son. I argue that Facebook comes to serve two critical functions for users, as both the site upon which life narratives are performed and organised, and also the site through which the variously public and private disclosures that constitute a persona are recalled and reflected upon. In setting out these arguments, I divide this paper into three parts: first, a description and reflection upon my own experience of the look back video; second, a consideration of critical moments selected for inclusion in the look back videos by algorithm as persona; and third, a discussion of death and memorialisation, as a sharp example of the significance of the digital traces we leave behind. The Look Back Video Gentle piano music rises as the “camera” pans across an assortment of photos. The flute joins the piano, and you are reminded that you started your Facebook journey in 2006. Here is your first profile picture—you with your arm around one of your good mates when you were twenty years old. Faster now, and here are “your first moments,” presented as images you have shared: March 2008, some of your closest friends who you met during your undergraduate studies, standing around sharing a drink; April 2008, a photo of a friend eating a biscuit, mid-conversation (she’d hate this one); and one last photo from April 2008, the biscuit-eating friend’s ex-boyfriend looking coy (you no longer speak to him, but he is still on your Friends list). Now enter the violins, seventeen seconds in. Things are getting nostalgic. Here are “your most liked posts”: July 2012, “thesis submitted for examination, yo” (46 likes); November 2012, “Trust me, I’m a Doctor… of Philosophy” (98 likes); February 2013, a mess of text announcing that you’ve found a job and you’ll be leaving your hometown (106 likes). Thirty-five seconds in now, and the pace of the music changes—look how far you have come. Here are some photos you have shared: December 2008, you at a bowling alley with your arm around one of your best friends who now lives overseas; October 2009, friends trying to sleep on your couch, being disturbed by the flash of your camera; June 2010, a family shot at your mother’s birthday. The pace quickens now, as we move into the final quarter of the video: September 2010, you on the beach with friends visiting from overseas; October 2011, you with some people you met in Canada whose names you don’t recall; (images now moving faster and faster) November 2011, ice skating with friends; March 2012, a wedding in Hawaii where you were the best man; December 2012, celebrating the conferral of your PhD with two colleagues; and finally July 2013, farewelling colleagues at a going away party. In the final ten seconds, the music reaches its crescendo and the camera pans backwards to reveal a bigger collage of photos of you and your nearest and dearest. Facebook’s trademark “thumbs up”/like symbol signals the end of the retrospective, looking back on the critical moments from the last eight and a half years of your life. Underneath the video, as if signing off a card accompanying a birthday present, is “Mark” (Zuckerberg, Facebook CEO, in a faux hand-written font) “and the Facebook Team.” Facebook is you, the note seems to imply; for our anniversary, we present you back to yourself (see fig. 1). On 4 February 2014, the look back video feature was made available to all Facebook users. Some 200 million watched their videos, and more than 50% shared them with their networks (Spiridonov and Bandaru). In other words, around 100 million Facebook users held up their own individually generated look back videos as a record of the persona they had crafted through the site, and shared that persona retrospective with their networks. The videos work in the same way that television news programs piece together memorial clips for celebrities who have passed away, blending emotive music with visuals that conjure up memories and reflections. The first point of difference is that Facebook’s look back videos were intended for the living (although this function shifted as I will explain in a case study towards the end of this piece) to reflect on their own personas presented through the site, and then (about half the time) shared with their networks. The second difference is the technical, automated process of piecing together, rendering, storing, and streaming these videos on a large scale. Spiridonov and Bandaru, two Facebook engineers writing on the site’s Engineering Blog, described the rapid development and rollout of the videos. They explain the enormous pool of technical resources and human capital that were brought to bear on the project, including thirty teams across the company, in just 25 days. They end their explanatory post with an homage to “the things [they] love about Facebook culture” that the project represented for them, including “helping hundreds of millions of people connect with those who are important to them” (Spiridonov and Bandaru). The look back videos also serve a deeper purpose that isn’t addressed explicitly in any explanatory notes or press releases: to demonstrate the great depth of disclosures users make and are implicated in by others on the site. In a one-minute look back video, these disclosures come to serve as the very digital traces that Bowker was interested in, forming a longitudinal record of the persona. Algorithms and Critical Moments Although the explanatory post by Spiridonov and Bandaru did not go into details, the algorithm that determines which photos and status updates go into the look back videos appears to consider the quantity of likes and (potentially) comments on posts, while also seeking to sample disclosures made across the user’s time on the site. The latter consideration works to reinforce the perception of the longitudinal nature of the site’s memory, and the extent to which the life of the user has become entangled with, enmeshed in, and mediated through Facebook. Through the logic of the look back algorithm, critical moments in the user’s life course—those experiences that mark out narratives of growing up—become measured not in terms of their value for individuals, but instead through a quantitative metric of “likes.” While after the initial release of the look back feature, Facebook did provide users with the functionality to alter their videos with some limited control over which images could be featured, the default was determined by the algorithm. Social network sites have come to serve as spaces for reflexive identity work, for the development of personas for young people (boyd; Livingstone; Hodkinson and Lincoln; Lincoln; Robards). The transition towards adulthood is punctuated and shaped by “critical moments” (Thomson et al.) such as moving out of home, dropping out of school, entering a relationship, learning to drive, a death in the family, going clubbing for the first time, and so on. In Giddens’ terms, the “fateful moment” (from which Thomson et al. borrow in conceptualising the critical moment), is “highly consequential for a person’s destiny” (121), and should be understood as distinct from but certainly affecting the inconsequential goings-on of daily life. When these critical moments are articulated and made visible on social network sit
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it