Raw Memory, Ephemerality of the Eternal in WhatsApp by Natalia Stanusch

Hauntology of Metadata

WhatsApp presents itself to us as a service that belongs to us. The sent images, the videos, the messages are encrypted for my privacy because they belong to me. But a single glitch can reveal the WhatsApp’s gaze lingering behind its interface.

Instead of seeing WhatsApp, I can see WhatsApp-me, a single byte of a user in a corporate web of billions. WhatsApp-me has little to no legal right over her messages as property, and no technical control over their existence if the app makes my archives disappear. WhatsApp-me can only blindly scroll through a gallery of black-out polaroid frames the moment all her multimedia chat history (films, pictures, gifs, stickers, voice messages) is gone. WhatsApp-me continuously falls into a glaring gap between the reassuring impression digital platforms give us of eternal existence.

A single glitch in a simple data transfer, and hours of weeks of months of photos taken and voices captured, the most tangible elements of chat, are gone. Only the metadata and text-content remained. Instead, scrolling through my WhatsApp chat history is like walking through an empty museum where all the artworks are gone and only the labels remain. But the aesthetics of void was not what haunted me. What haunted me was the notion of something being taken away from me.

The nature of any communications medium is that it embodies some fragments of communication process but restrains or eliminates others. Communication online, on such medium as WhatsApp, facilities two different temporalities: the temporality of a call and the temporality of a chat. The temporality of a video or a voice call takes place in a sphere of presence that stimulates physical conversation. Unless recorded or screenshotted, the call leaves almost no track of its content on the app besides metadata: date, duration, users involved. The call allows the content of the conversation to be saved (or forgotten) by relying on memory of the human actors involved. The temporality of chatting, as any electricity-based transmitter of information, annihilates the physical distance just as the notion of time and space. But, paradoxically, it fuses instant and suspense.

A significant part of forming relationships happens through a highly mediated artificial environment, which influences the way we remember, store and erase the history of a relationship. Many people do not dwell upon the consequences of transferring our sociality into rigid forms dictated by the medium’s properties. But the aesthetics and properties of WhatsApp facilitate several perceptual changes in how we conceive and relate to interpersonal memories. WhatsApp encourages a quick receive-send approach, where efficiency is amplified by the reassurance of repetition: re-reading a message, re-listening to a voice message, re-seeing a picture.

On the bodily level, all one needs to do is to indifferently (or frantically) scroll up or down the screen. The sheer gesture of scrolling already situates the body in relation to memory. A gesture of swiping is like caressing the elements on the screen.[1] But while a finger sweeping through WhatsApp archive could be petting the pictures and voice messages with tenderness and longing, it also implies an infinitesimal expression of boredom, speed, impatience, and indifference.[2] Memories in the mind are given away to the cloud. Human memory becomes device memory. The warmth of touch is replaced by the coldness of screen.

On WhatsApp, the past is chronological and equal. The episodes of silence are not indicated in any graphic way, no message is more noticeable than another. The antihierarchical structure of WhatsApp memory is somewhat blank in that one cannot search content based on date. But the context of conversation can be retrieved through search option which allows for browsing through the past, searching for specific memories (keywords) which appear in reverse chorological order (from the most recent to the oldest). The chat is navigated through the keyword search bar, whereas any media (voice messages, photos, films) are stored in a multimedia box under the media bar. The media bar seemingly despises the ontology of its content; there is no differentiation in WhatsApp ‘multimedia’ box, where pictures, films, and audio are stored altogether, with no filter option. WhatsApp mirrors the behavior of the Internet; as Hito Steyerl says, ‘[Internet] is a form of life (and death) that contains, sublates, and archives all previous forms of media.’[3]

The layout of WhatsApp favors the present over the past. It is designed to first show the newest content, forcing a bottom (new) – top (old) order of viewing. New messages appear on the bottom, and move ‘upwards’ the screen into the past. As if all those messages once sent, received and seen, could provide only a one-off emotional hit: their acting power is deadened the moment they are sucked up under the top bar of the screen, redirected into the past of WhatsApp, turned into history, nonimportant, semi-forgotten. Is this bottom-top mode of viewing a joke on sayings such as ‘off the top of one’s mind’ or ‘on the bottom of one’s heart’? But the bottom-top order is ricked in WhatsApp’s Starred Messages tab. Here appear both multimedia and text messages which were starred by the user as special, and are displayed in top-bottom order. Despite providing this seemingly never-ending archive of past friendships and loves, the app is not there to truly manage it. There is no memory curation, no narrative, only the raw, swipe-able record.

Do Not Be Concerned

WhatsApp serves as a reassurance that the memories stored there are everlasting. One can go back two years back in a chat history to quote a message or resend a picture. Accessibility across time and space: that’s what WhatsApp promises. But the ownership of the content I send and receive on WhatsApp is an illusion of ownership carefully cultivated by WhatsApp and its owner, Meta(/Facebook). WhatsApp had always been a free service economically justified by data gathering from users rather than from fees. The exchange of data manifests the omnipresent commodification of sustaining and extending social interactions online.[4] Since its acquisition in 2014, all the data gathered by WhatsApp is available to Meta(/Facebook), which has been reluctant to reveal exactly which data or how much data it gathers from WhatsApp or what purposes it is used for. Even if we don’t have any precise accusations to make, we should be concerned about the extent of our ignorance, or conformism; as Kylie Jarrett explains, ‘to accept the gifts of the platform is also to accept a social relationship in which the site retains all power.’[5]

The seemingly eternal archival nature is neither deliberate from a commercial nor an engineering point of view. In its never-ending aim to collect data, the archival infrastructure of WhatsApp is a side effect; the memories are saved but not safe. Data matters in its temporality because the outcomes of almost real-time data-prediction models and algorithmic categories are re-calculated and re-calibrated as new pieces of data arrive.[6] The human real-life experience conflicts with the temporality of algorithmic prediction and data collection in its quest for usefulness and measurability.[7] The preference for temporality and presence is reflected on the surface of the interface and deep inside the infrastructural logic of surveillance capitalism and attention economy. The storage aspect is neither a historical nor archival element, but a side effect of data mining and profiling.

…And How Does It Make You Feel?

What was eerie about the WhatsApp glitch that haunted me is that the data did not disappear entirely. Rather, metadata such as the kind of file that was sent, and when it was sent, was visible. It was only the content that had gone missing. What I had were the markers of the missing data, labels without bottles, the presence of absence. The chat history was also still there. The text messages were still there, quiet reminiscences of hours over years of friendship. They were there, still and omnipresent, aggressively bugging in their passivity. Like Tantalus in Tartarus with the dangling fruit just beyond his reach, I saw empty boxes that had once contained tangible, clickable memories. They were there to be clicked on, but I couldn’t click on them, and if I did nothing would happen. A selfie, a photo, a voice message, a short video, a meme (may gods be merciful to all those lost memes) were no more. They were replaced by tesserae of metadata. Or perhaps it was as if WhatsApp was playing the part of a comforter at a funeral, every empty box seemingly saying ‘we recognize your loss.’

The options of sending voice messages and pictures express the longing for the missing face-to-face cues. In WhatsApp chat, the touchable laughter or sob is resurrected each time a voice message is played back or a photo is clicked. These memory objects have affective power. But the affective power is still present once the objects are gone. That’s when data begins to haunt. The image is gone, the voice message is glitched, but their affective power is whispering through the void. Michael Peter Schofield argues, ‘the real link to the past, or even the authenticity of the associated memory is not as important as affect when it comes to haunting – that the ghosts must resonate with us somehow, with some prior experience, a feeling of loss rather than any specific meaning.’[8] But a single image is nothing for the app in comparison to the metadata attached to it; what constitutes the memory for the human, is a noise for the app. To separate a signal from the noise, which is an issue Hito Steyerl explores, while a necessity in large data sets, is ultimately an execution of power,[9] and in this deal, the cards are played out by the app. To borrow from Hito Steyerl again, ‘social relations are distilled as contact metadata, relational graphs, infection-spread maps, or just a heap of fake news.’[10]

Maybe it’s a part of the anxiety of the digital, as if having the access to these voice messages and photos was necessary to bring out the feelings those media evoked originally. The missing zero-one sensorial clues, photos and voice messages, became signs for temporal depth, like shadows creeping into the digital flatness. As if these missing puzzles were indispensable to access the same emotion that burnt while the message was still marked as ‘unread’ on my screen. If I had the power to play back the voice messages, to see the stickers, memes, and selfies, I would have had the power of evoking within me the same emotion that is brought while recalling a hug, or an exchange of gazes unmediated by any screen other than the screen of our own eyes and delusions. Amongst the everlasting void, something ephemeral is missing; the text is not enough. Or, maybe, simply, the digital has never been enough. Now I saw it: WhatsApp had the consent over my memory which I did not remember giving away.


[1] Martine Beugnet. ‘Touch and See? Regarding Images in the Era of the Interface.’ InMedia 8.1 (2020): 7, doi:10.4000/inmedia.2102.

[2] Ibid.

[3] Hito Steyerl, ‘Too Much World: Is the Internet Dead?’ e-Flux, 49 (2013), https://www.e-flux.com/journal/49/60004/too-much-world-is-the-internet-dead/.

[4] Kylie Jarrett, “Let’s Express Our Friendship by Sending Each Other Funny Links instead of Actually Talking”: Gifts, Commodities, and Social Reproduction in Facebook’, in Ken Hillis, Susanna Paasonen, and Michael Petit, Networked Affect, Cambridge, MA: The MIT Press, 2015, p. 204.

[5] Ibid, p. 213.

[6] John Cheney-Lippold, We Are Data: Algorithms and the Making of Our Digital Selves, New York, NY, New York University Press, 2019, p. 26.

[7] Ibid, p. 30; 92.

[8] Michael Peter Schofield, ‘Re-animating Ghosts: Materiality and Memory in Hauntological Appropriation.’ International Journal of Film and Media Arts 4.2 (2019), 25, doi:10.24140/ijfma.v4.n2.02.

[9] Hito Steyerl, ‘A Sea of Data: Pattern Recognition and Corporate Animism (Forked Version),’ in Clemens Apprich, Wendy Hui Kyong Chun, Florian Cramer, Pattern Discrimination, Lüneburg: meson press, p. 2, https://doi.org/10.25969/mediarep/12348.

[10] Ibid, p. 4.Na