How the Calculator Became a Computer

Lay out what we have gathered: switches (piece one), numbers held in switches (piece two), orders made of numbers (piece three). These alone do not yet give a machine-for-anything. Two more thoughts were needed — both born on paper, not in a workshop.

A machine on paper — one is enough

In 1936 a young English mathematician imagined an odd machine: it reads symbols on a long paper tape, and by a table of rules erases, writes, and steps left or right. That is all. He never built it. It lived in his head.

With this bare machine he proved something astonishing: change only the rule table, and this one machine can imitate any other calculating machine whatsoever. Give it an adder's table and it adds; give it a codebreaker's table and it breaks codes.

The same man pushed on to a deeper question — are there problems no machine can compute? — and a few years later, in the middle of a war, went off to build a codebreaking machine for real.

Putting the rule table inside

The first electronic calculating machines were far from universal. To change the job, people spent days pulling and re-plugging cables. The rule table lived outside the machine, as the wiring itself.

The thought that ended this came in the late 1940s. Orders are numbers anyway (piece three). Numbers can be stored inside the machine (piece two). Then store the orders in the same place as the data.

Changing the job stops being cable work and becomes merely reading in a different bundle of numbers — a moment, not days. The design was written up in a report bearing one mathematician's name, and nearly every computer since follows it.

A little further in

The price of universality

Orders and data living in the same place carries a price: something that entered pretending to be data can start acting as orders. That design is exactly why computer viruses are possible. The universal door opens for anyone.

Even so, the bargain paid. A machine with a fixed purpose becomes scrap when the purpose ends; a machine with none keeps becoming new machines as long as someone writes new tables. That is why a seventy-year-old design is still on duty.

The story of the first piece completes itself here. If half the history of the computer is making the switch smaller, the other half is the thought that took the purpose out of the machine.

Materials and blueprint are now assembled. So what actually happens inside this universal machine in the moment you tap the screen? The next piece follows a single click all the way down.

The question that remainsA machine that becomes anything when you swap the program — is 'the computer', then, an object or an agreement?