During that first year at Harvard, my father met a man who, had circumstances been ever-so-slightly different, would have altered the course of his professional life. That someone was an older graduate student named Howard Aiken.
Tall and handsome, with piercing, owl-like eyes, Aiken would become the force behind the world’s first fully operational general purpose computer, the so-called Harvard Mark I, a room-sized, fifty-foot colossus of camshaft and relays built by IBM to Aiken’s specifications at his urging, and that would be the centerpiece of a brand new degree program called “computer science.”
That Aiken and my father knew each other is doubtless. Most probably they met in a class on vacuum tube theory taught by E. L. Chaffee, an affable professor and formidable expert in electromagnetism and especially vacuum tubes. For the first half of the twentieth century, tubes were the prime component of all electronic devices. Until the 1950s, when much smaller transistors and other kinds of semiconductors started supplanting them, vacuum tubes were state-of-the-art, having themselves supplanted the equally reliable but much slower mechanical switches controlled by magnets called relays (“tick-tick things”).
By the time he and Aiken met my father was already an expert in vacuum tube theory, so much so that, before the end of that first spring term at Harvard, the Radio Corporation of America—RCA—would hire him away. As students of Professor Chaffee’s “Principles of Vacuum Tubes” class, Aiken and your father each had to work out complex theories of vacuum tube design. Since Aiken’s thesis involved the solutions of an enormous set of nonlinear differential equations, tedious calculations that had to be repeated many times over using various sets of closely-spaced numbers, he fantasized about linking up an array of commercial calculators and somehow configuring a system that would perform the endless sequences on them automatically, obviating his input. A computer.
It was with this fantasy born out of tedium and frustration that Aiken approached a series of powerful men, including Harvard’s President, James Bryant Conant, hoping to secure funding to design and build his dream computer. Instead he was told by Conant in no uncertain terms that if he persisted in devoting himself to computer fantasies rather than to electronics he’d have no future at Harvard. Not one to be dissuaded, Aiken took his brainchild to manufacturers of commercial calculating machines to see if they might be interested. When they weren’t, he turned to Thomas J. Watson, the founder of International Business Machines. Though aware that the resultant machine would in all likelihood have no commercial value, Watson agreed to have I.B.M. build it. It would be good publicity.
Why did Howard Aiken, a graduate student in electronics seeking to have a computer built to solve problems instigated by Professor Chaffee’s “Principles of Vacuum Tubes” class, knowing that tubes—not relays—were not only the state-of-the art technology, but far-better suited to the purpose, conspire to have the world’s first computer designed and built by a manufacturer of mechanical tabulating devices using relays, and not by an electronics firm with vacuum tubes? Because no electronics firm existed then with the resources to undertake so ambitious and unprofitable a venture. The money just wasn’t there. Thus what might have been a sleek humming instrument three times faster and half its size turned out to be a sluggish monstrous clicking mechanical device.
Had the world’s first computer been made with tubes, Paul J. Selgin would certainly have had a role in its development. As things went, however, the episode barely justifies the existence of this historical footnote composed by his middle-aged son on a device with which his papa had little—if anything—to do.