Before printed ballad sheets, before wax cylinders, public-house music lived only in the moment—unless the mug itself was listening. In 2030 a Norfolk metal-detectorist unearthed a dented pewter tankard whose base carried a 40-second snippet of a 1762 drinking song. Every rhythmic thump on the tavern table had flexed the metal microscopically; centuries of creep locked the strain pattern into the alloy’s crystal lattice. Using synchrotron white-beam Laue micro-diffraction and a visco-plastic inverse solver, engineers replayed the tune—complete with the off-beat chorus when the room joined in—turning a beer vessel into a self-writing musical medium.
Pewter (Sn 92 %, Sb 6 %, Cu 2 %) yields at room temperature under cyclic stress. Each table-strike (≈110 dB, 5 ms) imposed ±0.3 MPa hoop stress, nucleating dislocation cells whose density scales with loudness. Over 250 years grain-boundary diffusion pinned the cells, storing amplitude as a nano-strain grating sampled at musical tempos.
Reading starts by electropolishing a 5 × 5 mm coupon from the base. A 1 µm polychromatic X-ray beam maps Laue spot broadening every 200 nm; broader spots = higher dislocation density = louder note. A spiral scan yields a 1-D strain trace sampled at 96 kHz—sufficient for 4 kHz audio after correcting for pewter’s anelastic relaxation.
Clock recovery exploits the song’s meter. Drinking songs use 4/4 at 92 BPM; strain peaks repeat every 0.65 s. Cross-correlation with the 1762 broadside (kept at the British Library) aligns the trace to the score; one anomalous 7/8 bar matches a documented local jig, confirming authenticity.
Error correction uses musical redundancy. The chorus appears six times; stacking suppresses creep noise, boosting SNR by 12 dB. Weak signals—such as the 1 kHz fiddle E-string—emerge after median stacking, revealing ornamentation consistent with period field recordings.
Storage capacity is modest but culturally unique. One tankard base stores ~900 kB of strain data—across an estimated 30,000 pre-1800 pewter vessels still extant in UK museums, the potential archive is 27 GB of 18th-century pub music, predating the earliest folk recordings by 150 years.
Restoration is nominally destructive—the coupon is removed—but the void is cold-sprayed with atomised pewter, leaving the vessel acoustically and visually intact. Legal title follows UK Treasure Act: the object is private property; the audio, being immaterial, is released under CC-BY for community use.
For ethnomusicologists the lesson is clear: every dented tankard is a disc. Beneath the pewter’s tarnish and dents lies a dislocation lattice where the thumps of long-dead drinkers still beat time, waiting for the right X-ray pulse and the right creep kernel to step out of the metal and back into the tavern air.
