By Gardiner Tucker
I said about IBM’s research organization, when I joined the Watson lab at Columbia University in 1952, that it provided a wonderful degree of academic freedom, even though it wasn’t technically academic. That was the same spirit in which we started the Fellows program when I became director in 1963.
IBM Research had by the 1960s established itself at the forefront of a number of technical disciplines that we judged had the potential to lead to new hardware and software, as well as entire new fields of information systems. Recognizing our people for leading these breakthroughs was, at the time, through promotion to team leader or department manager.
What we needed was a way to encourage and reward individuals in a way that let them continue creative research, unencumbered by administrative duties. We also wanted to cultivate a way to encourage individual “gadflies” or “catalysts,” who could stimulate ideas in others, and help colleagues overcome bottlenecks.
This is why we decided to start the IBM Fellow program. We chose the name “fellow” by analogy with how universities recognized outstanding scholars.
One Research colleague who inspired me, and exemplified who a Fellow should be, was never actually named a Fellow. Phyllis Baxendale, who I knew from managing the research team in San Jose in the early 1960s, was programming a computer to do automatic abstracting, indexing and retrieval of documents. Her machine did this by extracting patterns of word usage from a set of documents, and comparing them with patterns from broader literature. This was pioneering work; an early harbinger, perhaps, of Watson?
Nathaniel Rochester is one of my favorite examples of a Fellow. Nat was already well known for designing the IBM 701 (the first general purpose, mass produced computer) and the first symbolic assembler. His CONCEPTOR project in 1955 attempted to simulate the brain’s neural network on an IBM 704 to more-efficiently detect patterns in a stream of data. Though it was not successful, we consulted John von Neumann about the architecture’s possibilities. He confirmed that it would have succeeded on a machine capable of simulating vastly more neurons than the 1,000 maximum on the 704.
Incidentally, at an annual meeting of the IBM board of advisors, I described the CONCEPTOR as part of my Research review. A shareholder reacted with “we’re in the business of business machines, not brain machines!” While the project was eventually dropped, it was this ability to think ahead of his time that earned Nat status as a Fellow in 1967.
There were outstanding individuals in many areas of Research. I picked these two because, while other industrial laboratories were also pursuing issues in device technologies and new electronics, IBM was unique in its equal emphasis on organization, programming and application methodologies – that is to say, on the revolution in information processing.
I am delighted to see that the Fellow program encompasses the entire company, and is still such an important part of IBM. When you hear the term “IBM Fellow,” think of a person who embodies a place with pioneering vision in an ever expanding field.
Note: Dr. Gardiner Tucker served as principal director of defense research and engineering in the U.S. Department of Defense, was named assistant secretary of defense in 1969, and assistant secretary general of NATO for defense support in 1973.