March 5, 2014
Irene Greif always thought she'd be a teacher. "For one thing," she told me, "I'd been told by my mother that it was good to be a teacher because you just worked the hours your kids were in school and you could come home." It had just always been the profession in the back of her mind, the default.
So then it must have been a bit of a shock when, after in 1975 becoming the first woman ever to receive a Ph.D. in computer science from MIT, Greif discovered that she didn't really enjoy teaching—she much preferred research. And so eventually she left teaching as a professor and did what she did best: studying, thinking, and figuring systems out. She founded a research field, computer-supported cooperative work, and has spent her life figuring out how to build better systems for humans to work together.
Greif recently retired from IBM, where she'd been since the mid-'90s, and is hoping to devote some time to encouraging young women to go into STEM fields and coaching them to stick with them—a twist on teaching that she does genuinely like.
I spoke with Greif recently about her experience as a young woman in a field with so few other women, about how things changed during the course of her career, and for what advice she wishes she'd had when she was first starting out. A lightly edited transcript of our conversation follows.
Why were you first drawn to computer science?
Well I got exposed to computers as early as high school. I was at Hunter College High School [in New York City] and we took a course in the college nextdoor, using an IBM 1401. So I started with punchcards and zeroes and ones and machine language and so on.
And what year would that have been?
That was probably my senior year in high school, so that would have been '64-'65.
I had always liked math. My mother had always liked math. She was an accountant, so that meant adding numbers. Over my career I learned about other notions of what math is and what a mathematician does. I liked logic problems, which I think probably drove me to computer science.
When I went to MIT it was early—they were just starting to define a computer science undergraduate degree. It was in engineering—which it still is now—and it actually had more of an engineering feel than I was comfortable with, and I kept slipping back and forth between math and the new computer science, and ended up getting a math degree instead.
There are some interesting things about that. There are a lot of kinds of math that are very close to computer science and I was able to choose some of those things and kind of piece together my computer-science major by making it have the feel I liked. But I think I also learned a lot about how different people think, by switching from the engineering mentality to math and back and forth, and it probably helped me in my career, in that it made me good at interdisciplinary kinds of work.
Can you talk a bit about what it was like being a woman at MIT during that time?
I think I was in the first big co-ed class. There had been women at MIT for ... forever, I'd have to check. But the big breakthrough for undergraduate women at MIT came in the 1960s, the mid-'60s, when Katharine McCormick donated a dormitory for women to MIT. [Editor's note: The money was actually donated at the tail end of the 1950s, but the dormitory took a few years to construct.] And that made a huge difference in whether parents would let their daughters go to the school.
In my class, that entered in '65, we had 50 women, in a class of 1,000. And that was big.
By now it's close to 50-50, so there's really been a huge change compared to what it had been.
I know you say that was "big," but that still seems quite small, at least compared to my own college experience. What was it like being on campus with so few other women around? What was it like socially?
Well, different people were different but, yes, there was a lot of working together and studying with the other women. There were certainly women who have stayed friends forever since then, and I have some friends from that time.
I always talk a lot about a woman, Candy Sidner, who was the third computer-science Ph.D. who, with a group of women, wrote a report, one of the first reports, on what it felt like to be a woman in the department and described the behaviors of men who were hard to deal with and so on. And I just always felt like, even being in the tiny cohort that she was in, having a couple of other women to talk to, made a huge difference in sorting out what's me and what's going on around me. And I never had that.
I remember, though, feeling that it was hard to find appropriate study groups. Because if you didn't find the right women doing the right courses in your dorm, finding a group of guys to work with was just, you know, for young women, laden with this is-this-dating-or-is-this-working-together kind of stuff. So it was hard.
I imagine it's easier now that there's a more even mix and you don't have to be—or try to be—the one girl in a little study group doing whatever the subject is you happen to be studying.
So, yes, I think we got very close to each other in the dorm, but I think it really was limiting and an issue.
Did you feel like the male undergraduates and male faculty respected your intelligence in a way that was equal to how they respected your male peers?
It's so hard to say. I mean, you know, we all had experiences of feeling like the professor kept looking directly at me to see if I nodded and got it. You did feel unusual and singled out in class.
I've certainly talked to people [women] in other fields who were told explicitly, you're going to waste your degree, or you shouldn't be doing this, or you should study X instead of Y. I never had that kind of experience at MIT.
So what was the next step for you, after finishing up your undergraduate work at MIT?
I stayed in graduate school at MIT, and so, in graduate school, my Master's and Ph.D., are in computer science. And I was, literally, the first woman to get a first computer science Ph.D. at MIT. There were other women who got computer-science degrees at other schools before me. So, you know, it's a mix of—it was early, and I'm old now—but also of which schools gave which degrees when. But I was among the earliest at all, and literally the first at MIT.
It is awesome, but I always talk a lot about a woman, Candy Sidner, who was the third computer-science Ph.D. who, with a group of women, wrote a report, one of the first reports, on what it felt like to be a woman in the department and described the behaviors of men who were hard to deal with and so on. And I just always felt like, even being in the tiny cohort that she was in, having a couple of other women to talk to, made a huge difference in sorting out what's me and what's going on around me. And I never had that.
I know you said that your mother was an accountant. Was your family very excited about your achievements in computer science? Were they supportive?
Oh yeah! They were thrilled! They were bragging! And they bragged that I was a girl among boys. They thought that was cool. [Laughs]
What came after you completed your Ph.D.?
Well, the interesting thing is that I sort of did it assuming that I would teach. For one thing, I'd been told by my mother that it was good to be a teacher because you just worked the hours your kids were in school and you could come home and all that stuff. So I had this teacher notion in the back of my mind, even though I went to get the Ph.D. and I wasn't going to be a K-12 teacher that worked that sort of hours.
My first job was at the University of Washington, Seattle. What I learned fairly soon was that I actually didn't really enjoy the mix of teaching and research, trying to balance the two. It just didn't work for me.
I think it also doesn't work for me to be a teacher. I give good single lectures but piecing together the story over a whole term ... I mean, maybe now I could do it, but I wasn't good that when I was young.
So it took a while. I went from Seattle back to MIT, but within a couple of years at MIT I switched to research staff and started focusing on just being a researcher. At the time I felt it was a little awkward. They have a kind of research track now, but then it still felt like the faculty were really the first class people there. But, you know, I have a whole history of always choosing marginal roles and in marginal subjects of research and so on for myself.
I did become a researcher, and slowly moved from these very mathematically oriented computer-science areas to much more people-oriented work—office automation and human-computer interface and so on. And that was progression over the years from '75 to the mid-'80s. And it was in the '80s that I started this research field, computer-supported cooperative work (CSCW), which is the major thing I've done in my life: getting a set of people together across disciplines who would look at social systems and computer systems at the same time.
Can you tell Atlantic readers a bit more about that work and why you were drawn to it?
When I came back to MIT, the hot topic in systems was distributed databases and ways of preserving consistency across copies of databases, and, at the same time, there was increasing interest in office automation and putting work online and helping people formalize business processes.
There was a point when I realized that a lot of the work we were doing on the database side actually resurfaced among people in offices. Even in the way people set up appointments or set up meetings in real life is like the way people were doing the database work. So if you are trying to set up a meeting with a lot of people, you ask all of them to pencil in some times until you find out what times everybody is available and then you go back in the second phase and lock in the time. So things like that, that were being done at a systems level were also being done at a people level, and it seemed more interesting to me to do it with the people.
I shifted to working in office automation, but actually learned pretty quickly that there was also a set of people starting to look at what happens in offices and realize that you could go too far with automation—that if you start to make processes a little too invisible, you break a lot of the important social things that actually make work happen. The simplest example of this is if somebody is walking a paper form around an organization trying to get signatures, one person might notice that somebody two signatures down is about to be on vacation, and you better rush ahead and find that person and come back to me later. If you've got that form moving around in a system and each person just sees what they have to do and then it goes to the next one, you lose the chance for that conversation, you lose the chance to adjust the process.
And that was really the beginning of the notion of who needed to be talking to each other among researchers, in order to really get things right around using computers to help people work together. It had to be a mix of the systems people who knew how to build cool distributed systems who knew how to build cool, distributed systems, and these ethnographers and anthropologists and sociologists and different kinds of designers—people who would be able to take into account these social issues—as they design system.
I feel like a lot of people would say that women tend to be more interested in collaboration, or the human side of systems. Do you think that your gender played a role in you finding this niche?
Probably! You know, it's hard to say—it's not as if the field is totally dominated by women, although we do see more women in some of these more socially facing settings, and it's interesting.
I was in a meeting with one of the most prominent women in the field recently, honoring somebody who was very big in all that distributed database stuff in the '80s, somebody who had been at Xerox PARC, MIT, and Microsoft and so on. It was his 70th birthday honoring him. And at this event, we don't think there were 10 percent women.
The more you get into the straight computer science, the more you do see that it's still male dominated. And certainly if you have the old-timers coming to an event, you'll see it skewed that way. It was quite a dramatic reminder of what things had been like in those early days.
On the other hand, I was having dinner with somebody the other night who was saying that he's so thrilled that the applicants to MIT computer science for jobs this year were 50 percent women, and so he feels like, oh my God, we finally solved the pipeline problem. And I don't think we've solved the problems, but it's a dramatic difference from years ago when people couldn't find any woman—would claim at least—they couldn't find the women to even interview.
How would you say, more specifically, things have changed during your career? What are your observations about the field today?
So computer science overall is pretty broad, and it goes from things that are very mathematical, and you'll probably see fewer women there, to the very social and user facing, where it's more balanced. Overall there are more women around in numbers, and numbers matter. As I was saying before, just having women to talk to, to get some perspective on what's going on around you, really matters. I think it becomes ... easier, for somebody moving through the system now.
In some ways, it's a set up—you know, depending on what you choose—a woman who is moving through the system with enough women to not notice, and then makes a choice that leaves her in her first job, or whatever, and is suddenly in a predominantly male setting, might have more trouble with that than some of us who knew we were getting into that. And any place you go where there's a long history behind you, even if they're making concerted efforts to change the numbers, you're just going to see imbalance, and it's slowly getting fixed.
When did you join IBM? What was the environment there like?
When I left academia I went to Lotus Development Corporation, so I joined IBM when we were acquired by them. I had actually worked at IBM once before as a summer intern, but this was really joining IBM, because of acquisition. It was kind of interesting—it was the first of many software acquisitions that IBM did—and it actually took, like, five years before we really had to act like IBMers. I was still part of Lotus and I ran a little research group in Lotus, but I did start—because I knew colleagues in research at IBM, I mean, it was a preeminent research institution in the world, and I had certainly known a lot of those people from conferences and so on—I started interacting with the research division, going to their strategy meetings and so on, from '95, when we were acquired, on.
I would say that IBM has overall a great track record on trying to do well by women, and it keeps getting voted best place to work for mothers and so on, but it was a shock to go to the research strategy meetings and look around and count the women to see how many were there, because it was much more of a hard-science atmosphere. At the time, hard sciences attracted more men than my CSCW community, but there were areas where women were making headway. My colleague Pat Selinger in 1999 was elected to the National Academy of Engineering for her work in databases. That was also the period of time when IBM was ramping up work-life balance programs to help working mothers.
By 2000 I became part of the research division, when Lotus had to be a regular division, my research group started reporting into the research division, and I reported at a senior enough level that I was able to be involved in supporting women and making sure women's voices were heard at review meetings. So I think I've helped some.
I have this other thing. As I've chosen these user-facing approaches, design-by-story and so on—I mean, I have been a bit of an outlier. While I think I've been respected, I also ... well, I gave a talk once about our approach to design at one of our staff meetings, and when I sat down, one of our male colleagues said to me, "That was very brave of you, Irene." So I kind of have been an outlier, not just by being a woman, but by choosing to do more user-facing work in pretty straight computer science. (And don't forget that IBM scientist Fran Allen became the first woman ever to win a Turing award, which was in 2007.)
If you could talk to your younger self, do you have any advice you wish you had had? Things that would have helped you when you were younger?
I had been a terribly shy little girl, and I always assumed that any issues I had, socially, were because it was me. And it would have been great to have been told, you know, it isn't you. But, you know, my mother and father would have been telling me that anyway, so I'm not sure where I could have had credible advice. Parents love you, so where do you get the advice you can believe? (I can’t help but add that I realize that I’m very lucky to be able to say this. My parents really thought I could do anything. I’m not sure all girls that get support at home.)
Just watching anyone else ahead of me might have made a difference. And it's because it is so important to have role models that I am still fanatical about making sure women are visible at by-invitation conferences and panel sessions—it's surprising how often people slip up on this even in fields where there are plenty of women. If you don't watch for it, you might inadvertently just have men on the program.
When you look back on your career and your younger self, what are you most proud of? What are the things you feel like you did really right?
I think it's a kind of mix of going with my gut, but mindfully.
I often will tell young women this when I'm doing career talks: I didn't have a planned career. And there's nothing wrong with that. Sure, I know people who know, every couple of years, what their next move should be. And I haven't done that, but I think I've taken the time at various points in my career to think about: Is this working for me? Is this right? Is it time for a change?
It's a being-true-to-yourself thing, and having some integrity about what you're doing, that I feel good about. I don't exactly how it came about that I would operate that way.
I think it served me well also, ending up in corporate labs where you're always trying to transfer ideas. So you have a brilliant idea, you write a great research paper about it, but how do you decide: Is this the moment to fight for transfer? So, product group's not ready, do I stick with it, or do I give up and go on to the next thing? You just always have to be reevaluating, rethinking. People are always asking me, how do you do tech transfer? And there isn't one way; it's what's right at that moment.
I think that's what I might say: This ability to be flexible and reflect and shift directions ... I think I've done right, perhaps inadvertently many times, but I've done it right over my life.
(Image via Flickr user parksdh)
March 5, 2014