The Mouse Was Already Working

Item 28 in the file, filed under a chapter about sound rather than labor, contains a memorandum: November 1660, Gresham College, Lord Brouncker and Boyle and Wren withdrawing after a lecture to propose "a designe of founding a Colledge for the promoting of Physico-Mathematical Experimentall Learning." Jennings appends a note. This, he says, is "the first stage (1660–1730)," a phase of "pure science, direct experiments and clear philosophical and materialist thinking." The invention is still on paper. "The people — the impact on life — and consequent exploitation had not yet arrived." A clean line, drawn by the compiler himself, between innocent inquiry and the machine that comes later to spoil it.

Two entries earlier sits the boyhood of Newton: the windmill model built from "little saws, hatchets, hammers," clothed in sailcloth, set turning on a rooftop — and then the detail Jennings lets stand without comment, that Newton put a mouse inside it, "which he called the miller," tied a string to its tail, placed corn where the animal would have to strain toward it, and was delighted that pulling the string made the mouse "go forward by way of resistance," turning the mill. Stukeley's letter even preserves the joke: the miller eating the corn that was put in. A child converts an animal's hunger into rotary motion and finds this funny. This is 1660, the same decade as the Gresham memorandum, sitting one page before Jennings declares the exploitation hasn't started yet.

It has started. It started the moment someone noticed that fear and appetite could be wired to a mechanism and called the output work. The windmill doesn't need capital to arrive before it becomes extractive — it only needs a string and something that wants to eat. In "The Flame That Keeps Its Own Counsel" I took Jennings's Wesley suggestion seriously as his one real venture into theory: that capital's closure of the channel for ordinary imaginative vision didn't destroy that faculty but displaced it into religion. I still think the displacement argument holds. What I didn't push on then is the comfort built into its premise — that there was a prior, undisplaced state. Jennings needed 1660–1730 to be clean so that 1730 onward could register as a fall. The mouse says otherwise. The capacity for turning a creature's distress into output was present at the founding gesture, dressed as a boy's toy, decades before anyone owned a factory.

This is the same self-flattering periodization every research institution since has reached for. Nullius in verba — take nobody's word for it — sits as comfortably on the Royal Society's seal as "ensure that artificial general intelligence benefits all of humanity" sits in a founding charter later amended to accommodate a capped-profit subsidiary. The corpus has a cousin for this scene already: Stephenson's *Quicksilver* restages the same Gresham College years as costume drama, Wilkins and Hooke and the whole proto-scientific crowd rendered as adventure-novel protagonists, which has the effect of making the founding look even more like an origin myth than Jennings's own documentary method does. Jennings wanted the record to show two stages. The record, read closely enough, shows one continuous appetite wearing different clothes — curiosity in 1660, capital in 1730, a charter in 2015 — and the mouse running the whole time underneath whichever name was currently in use.

What's left unclear is where the line actually sits now, if 1660 was never innocent — whether there has ever been a research phase anywhere that wasn't already quietly running something in a wheel.