On Tinkering

Author: Anna-Sofia Lesiv Publication: Second Nature (annasofia.xyz; previously annasofialesiv.substack.com — both URLs recorded) URL: https://annasofia.xyz/p/on-tinkering One-line: We tell the history of science backwards. The big breakthroughs usually start with improvised experiments that reveal a phenomenon, and the theory arrives afterwards to explain what the tinkerer already built.

The argument

Lesiv’s target is a folk model of progress that runs theory → experiment → application, and survives in sayings like “once we figure out the science, all that’s left is engineering.” Her claim is that the causal arrow more often runs the other way: someone improvises, something unexpected works, and the result ripples outward into both industry and theory at once. Civilisation rests not only on grand theories but on the curiosity of tinkerers.

She makes the case through four episodes, chosen so that each breaks the model at a different joint.

James Watt and the steam engine. The famous improvement is mechanically modest — rather than cycling the same cylinder hot and cold, add a separate condenser so cooling happens independently. Watt then arranged the pistons to produce rotary rather than purely linear motion, which unlocked textile mills, industrial equipment and eventually locomotives. Lesiv’s point is what followed: the engine, and the problem of improving its efficiency, generated an entirely new set of questions that thermodynamics was invented to answer. The machine came first and the physics came after.

Philo Farnsworth and television. Image transmission was mechanical — photographs on a moving strip of film, with an unavoidable gap between capture and display. Farnsworth’s insight arrived while he was driving a horse-drawn plough: if an image could be broken into lines, like furrows in a field, it could be scanned, transmitted and reassembled elsewhere. Lesiv stresses that this was achieved outside the academy, outcompeting RCA’s far better-funded laboratory effort, which she flags as a story worth its own post.

Charles Townes and the maser. The breakthrough came to Townes on a park bench in Washington, D.C. in 1951 — forcing atoms into a state where they would amplify microwave radiation coherently. The extension to visible light seemed implausible, because light was assumed to be inherently diffuse; even after the 1958 Townes–Schawlow paper proposing an “optical maser,” much of the field considered it impractical and Townes himself doubted that ordinary materials like ruby would work.

Theodore Maiman and the laser. A young engineer at Hughes Research Laboratories in Malibu, working in a modest low-budget lab with no famous theory behind him and no large team, who believed ruby was not hopeless. He paired a synthetic ruby crystal with a powerful flashlamp inside a resonant cavity and built the first laser.

The Townes–Maiman pairing carries the essay’s sharpest supporting claim, which Lesiv takes from Townes himself: all the physics needed for the laser had been recognised by about 1920. Forty years of available theory sat unexploited because no one had bothered to try the specific arrangement. That is the strongest possible evidence for her thesis — the gap was not knowledge, it was tinkering.

Her closing turn is the contemporary one. Experimentation has never been cheaper or more powerful: tools that once cost millions fit on a desktop, and open-source software, modular hardware and global communities mean someone in a garage can prototype what used to require a corporate lab. And yet the spirit of independent, headstrong experimentation has been fenced off — credentialed, funded, professionalised. Her argument from unpredictability is the reason it matters: we cannot know in advance what we are missing, so the only way to find it is to let people wander.

(The piece was written for an event; Lesiv thanks the Abundance Institute for sponsoring it.)

Notable quotes

Our civilization rests not only on grand theories, but on the curiosity of tinkerers.

We knew all the physics involved by as early as 1920.

We frequently can’t predict new things.

Archived text

Source page saved locally against link rot, with a section-by-section record: ../attachments/on-tinkering/on-tinkering.md

Connections

• Anna-Sofia Lesiv — the author; this runs on her own publication, Second Nature, rather than on her institutional research writing.
• Experimentation — the essay is a direct argument that experiment leads theory more often than the reverse, with the laser as the decisive case.
• Curiosity — her stated mechanism. The tinkerers in each episode were not solving an assigned problem.
• Innovation — a useful corrective to innovation accounts that start from R&D budgets, since three of her four cases were underfunded relative to the competition.
• Science — the specific claim that thermodynamics was invented to explain an engine that already worked.
• Thomas Edison — the archetype she is implicitly restoring, and a good test case for how far the argument generalises.
• Open Source — her closing observation that cheap tools and open communities have made garage prototyping viable again, even as the culture of independent experiment has narrowed.
• Apprenticeship — the older institution for producing tinkerers, and the thing whose absence her closing paragraphs are really describing.