The Granite Slab and the Flowing Stream: Contrasting Views on Preserving Social Media

Every day, a sprawling, cacophonous piece of our collective history writes itself across the servers of a handful of tech corporations. It’s on Twitter (or X), Facebook, and Reddit that political movements are born, cultural moments are meme-ified, and personal lives are documented. Preserving this digital commons is one of the most urgent challenges for archivists today, but there is no consensus on how to do it. Two distinct philosophies have emerged, and they stand in stark opposition: one seeks to create a definitive, granite-like record, while the other aims to capture the fluid, living context of the stream itself.

The first approach, which I think of as the "Granite Slab" method, is epitomized by the Internet Archive’s Twitter collections and similar initiatives by national libraries. The goal here is monumental: to capture a complete, verifiable snapshot of a platform at a specific point in time. Think of the archiving of the entire Twitter stream for the day of a national election or a major world event. This data is often collected via API, structured, and stored with the intent of being a citable, academic-grade resource. It’s the digital equivalent of chiseling a transcript of events into stone—authoritative, permanent-feeling, and incredibly valuable for future historians who need to know, with precision, what was said and when.

Yet, for all its authority, the Granite Slab has a fundamental weakness. It captures the words but often misses the music. It freezes a moment, but in doing so, it drains the life from the stream. The relentless, algorithmic timeline, the subtle shifts in a Like count, the real-time replies that change the meaning of an original post—these are the contextual currents that give social media its texture and its truth. A static archive cannot replicate the experience of watching a news story develop in real-time, with all the confusion, correction, and communal sense-making that entails.

This is where the second philosophy, the "Flowing Stream" approach, finds its purpose. Instead of periodic snapshots, this method seeks to document the experience of the platform over time. It’s less about preserving a perfect copy of the data and more about preserving the user’s perspective. Tools like the Webrecorder project allow individuals or institutions to create high-fidelity, interactive captures of their social media feeds, complete with the interface, the scrolling dynamics, and the ad-hoc curation of their personal algorithm. It’s an attempt to bottle the stream, not just the water.

The Stream approach acknowledges that on social media, the medium is an inseparable part of the message. The outrage fueled by a recommendation algorithm, the echo chamber of a personalized feed, the visual design that prioritizes certain content—these are critical to understanding the cultural impact of these platforms. A future historian studying the Flowing Stream archive would understand not just what people were saying, but how the platform itself shaped the conversation.

Neither approach is perfect. The Granite Slab risks being a sterile dataset, disconnected from the visceral reality of the lived experience. The Flowing Stream is inherently subjective and personal, making it difficult to assemble into a coherent, society-wide record. The ideal preservation strategy likely involves both: the Slab for the raw, verifiable facts, and the Stream for the rich, messy context. As we grapple with how to remember the digital agora, we must accept that its memory may need to be housed in two different kinds of vaults: one a library of stone, and the other, a museum of currents.

Notes & further reading

A few pages I came back to while writing this: