The Conservator's Palette: What Data Preservation Can Learn from Oil Painting Restoration
In a quiet, climate-controlled studio, a conservator works not with code or servers, but with solvents and scalpels. Their subject is a centuries-old oil painting, its surface clouded by yellowed varnish and its canvas weakened by time. This meticulous process of art restoration, seemingly a world apart from our digital concerns, offers a profound and unexpected framework for how we approach the preservation of our public records and data.
At first glance, the goals appear contradictory. A painting restorer seeks to reveal the original work by carefully removing the layers of history—the grime, the overpainting, the discolored coatings. A digital archivist, by contrast, strives to save every layer of history—the original file, its subsequent edits, the software that rendered it, the operating system that ran the software. One is an act of subtraction to return to a point of origin; the other is an act of addition to capture a continuum.
Yet, the core philosophy is the same: radical transparency. A master conservator would never simply paint over a damaged section. Every action is documented in meticulous reports, accompanied by high-resolution photographs and spectroscopic analysis. This creates a ‘treatment history’ that is inseparable from the artwork itself, allowing future experts to understand what was done and, crucially, to reverse it if better techniques emerge.
This is the lesson we must borrow. Our digital preservation efforts often focus on freezing a perfect, ‘original’ state. But data is rarely a static masterpiece; it is a living document, a public record amended with new ordinances, a dataset corrected with new entries. We should aim not for a single, pristine snapshot, but for a richly-layered and thoroughly documented provenance. Every change, every migration to a new format, every fix applied to a corrupted file should be logged not as a secret repair, but as a visible, reversible, and well-documented part of the object’s history.
The conservator’s most sacred rule is reversibility. Any material added must be removable without harming the original. In our world, this translates to avoiding proprietary, ‘black box’ solutions that lock data into a system that may itself become obsolete. It demands we use open formats and standards, ensuring that our preservation actions themselves do not become the very agents of entropy we seek to defeat.
Viewing our work through this lens reframes the entire endeavor. We are not just technicians backing up bits. We are conservators of public memory. Our responsibility is to treat each dataset, each archived website, each digital record with the same reverence, transparency, and foresight as a restorer treating a Rembrandt. We must ensure that the choices we make today—our digital solvents and scalpels—do not cloud the view for those who will come after us, armed with tools and knowledge we cannot yet imagine.
Notes & further reading
A few pages I came back to while writing this: