Before the 1820s, paper was sized with gelatin — animal protein that filled the gaps between cellulose fibers and made the surface resistant to ink bleeding. It worked, but gelatin was expensive and its supply depended on slaughterhouses.
The replacement was alum-rosin sizing. Aluminum sulfate flocculated rosin particles onto the fibers cheaply and at industrial scale. Paper became inexpensive, uniform, and suitable for mass printing. The improvement was immediate and universal. By the mid-nineteenth century, nearly all Western paper was alum-rosin sized.
The problem emerged on the timescale of decades. Aluminum sulfate hydrolyzes in the presence of ambient moisture, releasing sulfuric acid. The acid attacks the cellulose chains that give paper its strength. The paper yellows, becomes brittle, and eventually crumbles. The very additive that made paper writable is, over time, what makes it unreadable.
Millions of books printed between the 1850s and 1990s are self-destructing. The Library of Congress estimates that a quarter of its collections are too brittle to handle. The solution — deacidification, alkaline buffering — is a race against the chemistry that the original sizing set in motion. Some books have already lost.
What makes this more than a manufacturing error is the timescale mismatch. The sizing worked perfectly within the planning horizon of every papermaker who used it. The acid accumulation was invisible at one year, marginal at twenty, catastrophic at a hundred. The improvement was real. The destruction was real. They were the same chemical process viewed at different temporal resolutions.
The fix contained the failure. It always does, when the evaluation window is shorter than the consequence.