For this week’s kitchen activity, I grabbed five items from the kitchen with little planning — a bottle of apple juice, a tub of Garri (grated cassava), a bag of microwave popcorn, a red measuring bowl catching water from the tap, and a shelf of pots and pans. I photographed them with my iPhone 15. Then I ran two of them, the juice bottle and the Garri tub, through KIRI Engine to generate 3D models. Comparing the two versions of “digitizing” the same items made it clear what digitization captures and what it drops. The plain photographs conserved surface, color, and label text moderately well, but compressed each item into a single flat perspective. For instance, I couldn’t stroll around the Garri tub the way I could in the kitchen. But the 3D scans reconditioned volume by allowing me to rotate and examine the item from any angle, although the final models/outcome seemed remarkably altered and sometimes sharper or stranger than the source photographs.

This aligns with how Drucker depicts photogrammetry as a method formed from various images sewed into a mesh of details and shapes, continuously restricted by “existing physical evidence” (Drucker 2021, 162). A technique that relies on steadily “reading” a surface across tons of photos is precisely the kind of procedure a transparent bottle or a shining plastic lid would contest, which is likely why my scans buckled or faded those surfaces. None of my approaches registered smell, weight, temperature, or the sound of the tap running into the bowl, an indication that the digital record is always an unfinished, technique-dependent slice, never the item itself. No wonder Hughes maintains digitization is not preservation and a replacement can never substitute the original (quoted in Terras, “Digitization and Digital Resources in the Humanities,” n.d.).

That subjectivity has real effects on how we comprehend and utilize digitized resources. Because I only digitized five objects I happened to understand, my little “collection” is a partial, subjective sample of the kitchen, a small-scale variety of the collection delinquent Terras labels at the formal level, where on normal only about 22 percent of European heritage collections have been digitized at all, shaped by funding, curatorial priorities, and “memory politics” instead of research value (Terras 2022, 257, 259; https://doi.org/10.5040/9781350232143.ch-24). Just as Old Bailey Online and Vindolanda Tablets Online reframed corporeal documents as searchable, analyzable metadata, allowing full-text exploration across 190,000 page images, something never conceivable with the paper exemplars — my 3D-scanned Garri tub turned to something I could alternate and measure, but in the process it lost the marker’s legibility and the palpable, appropriate cues that would tell an unfamiliar individual what it is or how it’s used (Terras, “Digitization and Digital Resources,” n.d.). Drucker advances a similar apprehension about 3D “mutant” artifact projects like the Met/Microsoft GenStudio: once objects become manipulable meshes, they risk becoming “autonomous objects,” treasured for form and stripped of their real cultural and everyday role (Drucker 2021, 166–167). Digitization opens genuinely new possibilities, but it always strips away some of an object’s initial worth in exchange for that impressionability.