Making Small Talk
They weren’t necessarily trying to change the world. “We were just having fun,” Goldberg says. But they changed it anyway.
Goldberg (A.B. 1967, Sc.D.Hon. ’14) first became interested in computers as a mathematics major at U-M. After her junior year, she spent the summer working for IBM, teaching herself how to program on large physical boards that were operated like an early telephone switchboard. From there, she got her Ph.D. at the University of Chicago, with a focus on computers in education. At the suggestion of a faculty member, she went out west to Stanford, where she conducted research on computer-aided instruction.
When Goldberg began her work, large mainframe machines were the norm. These bulky and expensive computers were found in universities, research facilities, and businesses, where interested users paid by the hour for time-sharing to perform complex calculations and data management. The idea of a personal computer for the common person, let alone a child, was virtually unthinkable to most people.
Silicon Valley, it turned out, was the perfect place for a young computer scientist like Goldberg, with its wealth of big ideas and an ever-growing number of high-tech companies. This included the legendary Xerox Palo Alto Research Center (PARC), an innovation-focused offshoot of the Xerox company tasked with creating the tools that would be used in the office of the future.
Goldberg was hired at Xerox PARC in 1973, and along with Alan Kay, Dan Ingalls, and others, got to work developing Smalltalk, an object-oriented computer programming language. This type of language dealt with interacting objects, each object a set of actions combined with the data needed to carry out those actions, which they hoped could make it accessible to ordinary people—even intuitive enough, says Goldberg, for children.
“From a programming point of view, our goal was a language in which you can simulate how you understand real-world phenomena,” she says.
One of the earliest accessibility problems Goldberg’s team tackled was the computer’s user interface. At the time, computer screens displayed mostly in black and white using lines of text with no manipulable graphics. Smalltalk changed all that. Their new color graphics interface required the use of yet another little-known accessory, the computer mouse, in applications incorporating direct manipulation of displayed information.
Another concept introduced by Smalltalk is almost impossible to imagine living without today. “The biggest impact of Smalltalk was an accessible editing and testing environment for programming,” says Goldberg. “Ours was the first multiple-window system.” In other words, it was the first system where you could copy text from one window and paste it into another, or work on two different tasks at a time without losing what you were doing in either.
Smalltalk also offered users rapid prototyping, a process that allowed programmers to test parts of their code (actually object definitions) working incrementally, rather than completing the entire product and then testing, when problems could be more costly to find and correct.
Remarkably, the Xerox team also envisioned Smalltalk being used on a portable personal computer they called the Dynabook—a device whose prototype looked uncannily like a modern iPad, including multimedia applications.
(ABOVE LEFT) PARC in Palo Alto, California, served as the creative foil to Xerox's more buttoned-up corporate headquarters on the East Coast. (BELOW LEFT) The Dynabook, originally conceived as a “personal computer for children of all ages,” the Dynabook was imagined as a 9"x12" screen with a connected keyboard. Photos courtesy of PARC, a Xerox company (RIGHT) “When we were developing Smalltalk,” Goldberg says, “there were no personal computers. The idea that you could carry one around in your back pocket was crazy talk.” Photo by Ann E. Yow-Dyson/Getty Images.
“The goal was to have a programmable small computer you could carry around with you," says Goldberg. “The primary thing we were interested in was: How can people who are not professionals write their own programs? And why would they want to?”
Smalltalk caused a sensation among programmers. In 1981, the microcomputing magazine Byte dedicated an entire issue to the language and its new ways of conceptualizing software. Many of the ideas introduced by Smalltalk were integrated into later systems, such as Java and Apple’s Objective-C.
In 1988, after testing Smalltalk with both children and adults, Goldberg decided to spin off her own venture, ParcPlace Systems, which focused on marketing Smalltalk-80 to the outside world. Under Goldberg’s leadership, Smalltalk was introduced to a multitude of businesses worldwide, from telecommunications companies to financial institutions, to help manage databases, analytic systems, and more.
“Once, at a conference, Steve Jobs spoke and suggested no one would want to use object-oriented programming to build new code if you already had something that worked,” says Goldberg. “The next speaker was one of our clients—a car manufacturer—who said they were rewriting their (functioning) payroll system because, in negotiations with their union, they wanted to say yes more often to union demands but couldn’t because of the inflexibility of their current payroll software. Instead, they were working with us to make the consequences of change more predictable.”
While at ParcPlace, Goldberg noticed that one of the barriers facing software engineers was in their ability to manage large development projects, especially with team members in remote sites. Never one to shy away from an interesting problem, Goldberg founded a new company in 1997, Neometron—“new measures” in Latin—to address it.
“There weren’t network-based project management systems at that time,” says Goldberg. “So we tried to combine that goal with figuring out how people could work remotely from home, including communicating with one another.”
Neometron was based on ideas from Goldberg’s Xerox team, which, in the early ’80s, set up a satellite office in Portland, Oregon, to test their ideas about collaborative technology. Their work was one of the earliest collaborative systems, a development today’s growing legion of telecommuters is intimately familiar with.
Throughout the entirety of her career, Goldberg was a trailblazer, not only recognized as one of the top computer scientists in the field, but also as one of its few women. In addition to her many ventures, she served as the president of the Association for Computing Machinery (ACM) from 1984 to 1986, and received ACM’s Software Systems Award with Kay and Ingalls in 1987. In 1990, PC Magazine honored her with a lifetime achievement award.
Today, Goldberg is retired, but she is still active in the scientific community, and she maintains connections with other scientists around the globe through her membership on the science advisory board at the Heidelberg Institute for Theoretical Studies and the visiting committee for the University of Chicago Physical Sciences Division.
“I never really had a nine-to-five job,” she says. “But when you are retired, now it’s on you to ask what would give you joy in how you spend each day. More often than not, you want to be engaged with interesting people and still contribute to the world.”