In 1955, John Backus began adding to the team, and he plucked recruits from various sources. Backus visited MIT, a major center of computing research, which had a close relationship with IBM. He explained the FORTRAN project and asked if the university’s Digital Computer Laboratory might want to send someone to work on FORTRAN. MIT dispatched one of its star programmers, Sheldon Best. Backus was able to exploit the one flicker of interest he got from the industry, from United Aircraft, which lent Roy Nutt to the FORTRAN project. And Nutt was a real catch – an extraordinary programmer who could “execute” a program in his head, as a machine would, and then write error-free code with remarkable frequency. Nutt, Backus marveled, would walk straight to the keypunch when he felt he had solved some software riddle and, without notes, write a flawless program onto the punched cards. Inside IBM, Backus tapped people he spotted or who were recommended. He wanted people who were bright and seemed to have knack for programming. “There was nothing formal about it,” he recalled. “We added one person at a time, and it just sort of happened.”
The experience of the recruits was often slender. Robert Nelson, a former cryptographer for the State Department in Vienna, was a new employee hired by IBM to do the routine work of typing scientific documents. But Backus soon recognized Nelson’s technical talent. “He quickly became an outstanding programmer, absolutely crucial to the FORTRAN project,” Ziller observed.
Richard Goldberg had a Ph.D. in math from New York University, and he had intended to teach. A semester at Dartmouth College convinced Goldberg he was not meant for teaching, so he moved back to New York and got a job at IBM. “I didn’t know anything about computing,” he recalled. But Goldberg excelled in a three-month programming course, and he was sent along to Backus. Lois Haibt went to IBM straight from Vassar College, where she recalled being “good in math and science and terrible in the fuzzy subjects like English.” A scholarship student, she was more than “good” in math and sci-ence and her summer jobs were at Bell Labs. When she graduated, however, IBM lured her with a starting salary nearly twice the Bell Labs offer – $5,100 a year, which at the time seemed a lavish sum to her. “They told me it was a job programming computers,” Haibt said. “I had only a vague idea what that was. But I figured it must be something interesting and challenging, if they were going to pay me all that money.” In her programming class, she proved extremely adept and was assigned to the FORTRAN team.
For David Sayre, computing was at first merely a tool which later became a career. In the early 1950s, Sayre was a crystallographer doing biophysics research at the University of Pennsylvania. His research focused on the structure of hydrocarbon carcinogenic molecules. Even then, such study of crystal structures relied on computing assistance. Frustrated by the university computer, Sayre went looking for a faster machine. He had his eye on an IBM 701. He was told time on the machine cost $400 an hour, but that for his scientific research he could get a couple of hours free. The IBM officials had markets in mind, seeing work like Sayre’s as a way to make deep inroads into scientific computing. Sayre went to New York City, wrote his program line by line in octal notation and put it on the 701. He found the process intellectually satisfying and oddly compelling. “You entered a world that kind of ran the way it was supposed to, a world made for working out the logic of something,” Sayre recalled. “When you ran your program the expected thing happened, and if it didn’t there was a logical reason why it didn’t.” Sayre joined IBM, working on scientific applications but also going deeper into pure programming. He wrote a diagnostic program to help find bugs on the IBM 704, for example. So Backus “borrowed” Sayre for the FORTRAN project, and Backus had a way of holding onto the people he borrowed.
