We measured a range of objects, pre-industrial and machine-made, in areas not heavily subjected to wear: tabletop thicknesses, turning diameters, drawer-side thicknesses, etc. Using digital calipers, measurements were made at multiple points on each part and catalogued. The goal was not just to generate pages of numbers, but rather to establish some general guidelines – surface tolerances, variations in pre-industrial versus industrial pieces. We sought to define how irregular hand-prepped surfaces were, on average, and how this compares to those made using machines. To minimize the potentially confounding factor of warpage in our handmade test pieces, we also drew from a selection of Victorian and late Federal furniture to represent the machine-made pool. All wooden objects naturally fluctuate to some degree with the seasons, but after more than a century, any radical movement to which the wood was naturally predisposed should have already taken place. These late 19th- and early 20th-century pieces settle into a kind of low-stress stasis with no new, unpredictable warpage taking place. More recently built furniture might not work out these forces for years, and might therefore appear more “regular” than objects whose constituent parts have found their natural equilibrium.
Setting aside outliers for the moment (which I will discuss below), our findings were remarkably consistent. In brief, show surfaces that were intended to be regular on pre-industrial furniture maintained an average of 0.03" variation across the board, while machine-processed pieces averaged 0.01" variation. These results caused us to consider whether this 0.02" difference (half a millimeter, roughly the thickness of four pages of the magazine you hold in your hands) sufficiently characterizes the breadth of the handmade aesthetic, and defines what visually sets pre-industrial furniture apart.
–Michael Updegraff, excerpt from “In Pursuit of the Handmade Aesthetic,” in Issue Four