The experience curve originates in an 1899 psychology study of telegraph operators — whose learning curve had plateaus that Wright's 1936 industrial law smoothed away

The cost-per-doubling "experience curve" is usually traced to T.P. Wright's 1936 aircraft-manufacturing paper (see claim-wrights-law-cost-falls-per-cumulative-production-doubling). Nordhaus roots the underlying learning curve a generation earlier and in a different discipline: he attributes the original concept to "telegraph operators in W.L. Bryan and N. Harter," whose 1899 Studies on the Telegraphic Language (Psychological Review) charted how Morse operators acquired sending-and-receiving skill over months of practice. The curve was a fact about human skill acquisition before it was ever a fact about factory cost.

Its founding shape differed from the industrial version in a way that matters. Bryan & Harter's curves featured plateaus — extended flat stretches where measured skill stopped improving, followed by renewed jumps, which they read as stages of qualitative reorganization (mastering letters, then words, then whole phrases). Wright's 1936 law discarded that structure, fitting cost decline as a clean log-linear function of cumulative output with no plateaus at all. The smoothing is not merely cosmetic: a plateau is precisely where a naive extrapolation of the smooth curve would over-promise continued gains. The plateau itself is contested — Keller's 1958 "The phantom plateau" argued the phenomenon was a measurement artifact — so the note records that Bryan & Harter reported plateaus, not that plateaus are a settled feature of skill learning.

The origin sits beside the vault's telegraphy cluster: Bryan & Harter's 1899 Morse operators are the same occupational population, in the same decade, as the sufferers of telegraphist's cramp — one literature framing the telegraph key as a site of skill acquisition, the other as a site of occupational injury. It also connects to Nordhaus's broader critique: the same author who exposes the modern coefficient's fragility also points back to the curve's plateaued, psychological ancestor.