Dr. Radia Perlman
First, I’d like to thank the Anita Borg Institute for awarding me this honor, as well as the people who nominated me. As for a speech, I had a hard time deciding what to say. I’m used to speeches in which I can draw pictures of packet formats. And in this forum I should probably say something about being female, and how it affects someone in the technical world.
I would never have imagined myself in a job in the computer world. Although I was always good at math and science, I was never a “hands-on” type of person. I was intimidated by the guys who talked about how they’d built ham radios when they were 7. Although I was good at the logical world of math…(”this is the definition of a blotz, this is the definition of a glitz…prove all blotzes are glitzes”), I had no idea how a radio worked, and certainly would never take one apart.
I wound up programming as a part time job during college, in the MIT Logo lab. Logo is a language for teaching children programming. I designed and built a system for teaching all the concepts to children as young as 3. In my system, computer commands were tangible objects that you plugged together into a program. The project was exciting enough that, although I’d forgotten about it since it was so many years ago, people from the MIT media lab tracked me down a couple of years ago saying “You are the mother of tangible computing!”. I said “What’s tangible computing?”
Which brings me to a gender reference. To be painfully honest, I was a bit embarrassed by the whole project. I was just about the only female around, and I wanted to be taken seriously as a scientist. And my project involved all these cute little kids.
I was in the math department of MIT at the time. In addition to the special hardware involved in the tangible computing project, I was a student. First an undergraduate, then a graduate student. I was very shy and insecure. My perception was that everyone else in grad school was there because they were smart, whereas I just got in because I studied really hard. I couldn’t imagine doing original research. I was too afraid to talk to the professors because I’d be discovered as just an ordinary person, not the other kind of species capable of original research. Turns out everyone feels this way, but nobody tells this secret. But anyway, after I’d done all the coursework, and the only thing left was a thesis, an old friend invited me to join his group at BBN, designing and building network stuff. I found I really enjoy network protocols.
I see protocols in everyday life. Like at meetings, watching the various channel acquisition protocols of the meeting participants. The ones that just interrupt whenever they want, the ones that interrupt more loudly, the ones that will stop when someone interrupts them, the ones that raise their hands, and the ones that wait until there is silence before starting to speak. As you can imagine, with no standard for how to participate, the airtime is not shared fairly.
In the world of protocol design, my math background is very helpful. Math teaches you to think cleanly. Sometimes the really hands-on people just rush off and start doing things, attacking problems that are clearly unsolvable, instead of thinking about what would be the right problem to solve, and building something clean and simple.
Early in my career there was an intervendor meeting that happened regularly. The person running the meeting mentioned a routing problem that he wanted people to think about. I thought about it awhile, and when looking at it from the right direction, came up with a very simple, obvious solution. At the next meeting, I got on the agenda, and presented the solution.
Nobody asked any questions, and at the end of the meeting, the guy running the meeting again said “So, remember everyone, there’s this really important difficult routing problem I want you all to think about”, which was the problem I’d just presented the obvious solution to. As outrageous as that incident might sound, this was unfortunately typical of my experience. It’s always irked me that pompous people are so successful.
It’s common, when someone with an air of self-importance gives a talk and is totally incomprehensible, for people to think, “Wow. That must be a really smart person, because I couldn’t understand anything he said.” And people are impressed by complicated solutions. Whereas, when I speak or design things, the result is so simple and clear that people often do not realize the work and creativity that went into finding that simplicity, and they think of the result as obvious, and do not attribute it to me.
Anyway, after the meeting I was referring to, someone from Digital came up to me and said, “Thank you for your solution. We will use it in DECnet. And, if I may ask, are you happy professionally?” I said, “Yeah, sure why not?” And he said, “Well, you were just completely ignored. I never saw anything like it.” I said, “I’m used to being ignored.” He said, “Well, you have to come to Digital. You won’t be ignored there.”
It took a lot of courage for me to take that job. It was 20 miles away from the daycare that was on-site at BBN, and I had a 6-month old baby. But I am glad I did that. I wound up being the one to design the routing for DECnet. Although you don’t hear about DECnet anymore, the basic algorithms that I designed are in the protocols running in the Internet today.
Despite my work having such a profound effect on the field, people still did not particularly remember who was behind the designs. What I think really changed things for me was writing my first book, “Interconnections”. That book became the standard for learning thefield. If people have learned the computer networking field from my book, it doesn’t matter what I look like, or act like. They will take me seriously. I wound up taking a break from Digital and going back to grad school to get my PhD.
It was wonderful the second time. Doing a thesis was no longer intimidating. I knew the field better than anyone else. I didn’t need an advisor. I’d learned secrets like that it’s OK to ask questions in class, that you should start on a problem set the day it is assigned rather than the night before it is due. That there’s no need to panic for a test if you’ve been doing the work.
One really fond memory I have from that time was that I was at a playground with my daughter, who was about 7 at the time. My daughter was so proud of my being in grad school. She was playing with another little girl, and said to her, “My mommy is going to get a PhD from MIT!” The other little girl said, “My mommy already has one!”. Which was true…it was in chemistry, and we got to be friends.
My proposed thesis topic was designing a network that would work with malicious failures of trusted components. Routers that might lie in their routing messages, or flood the network with garbage, or perform the protocol correctly, but then fail to forward packets properly. This sounded like a hard, important problem, and luckily my thesis committee agreed it would be adequate for a PhD. Actually, one member said, “What? You don’t already have a PhD? With all the stuff you’ve done already? Your work to date already deserves a PhD. Whatever you’re happy with, I’ll sign.” Another member said, “This seems like a great opportunity for impossibility proofs” (since she assumed it would be impossible). The third member said, “Oh. This is so difficult and important that if you can have a network where two adjacent nodes can talk despite malicious behavior of some other node, that would be enough.”
As it turned out, looking at the problem from the right angle, the solution was simple. Rather than impossibility results, I had shown how to build a network that would have the property that two honest nodes could communicate provided at least one functional path connected them. And the approach was even practical in moderate sized networks. I was alarmed at how simple the answer was, and worried about whether I could get a PhD for something so simple. I asked someone if there was any minimum length for a thesis. He said, “Well…it either has to be long…or good.”
So I got my PhD, which was nice because although I seldom use a title, if someone obnoxiously insists on knowing whether I’m “Miss” or “Mrs.” I can say “Dr.”
Mostly these days I work on network security. Although my thesis is viewed as important work in the security community, I hadn’t really thought about the problems the mainstream network security community was working on. The way I got into that was that, seeing how my book changed my career, I was trying to encourage the security architect at Digital, who was humble and therefore underappreciated, to write a book. He finally said, “If you’ll be my coauthor”.
There is no better way to really learn a field than to write a book.
Back to the gender thing. Why are there so few women in the field? And what can we do to change it? Sometimes, especially when I am at standards meetings, it reminds me of the book “Lord of the Flies”. A bunch of adolescents with no adult supervision. The people who get to be the most powerful are politically adept, and vicious, rather than technically good. Women are tolerated if they are wide-eyes worshipers of the inner circle. Sometimes this same culture persists in companies. And I’d never work at one of them, at least, for long. Although there are some pompous, self-promoting women, and certainly a lot of men that are wonderful to work with and supportive of others, I’d like to think that the culture would change if there were more women. Also, an ideal team consists of people with different types of skills.
If people just relate to and hire and promote people that think and act like themselves, the technical output will suffer. But how can we change things? I am nervous about affirmative action, especially if less qualified women are promoted. It casts a cloud over achievements of women who have earned their achievements, and ultimately reinforces stereotypes of women as less capable. I got an email recently from a recruiter for a high tech company saying that they were very interested in me as a “female thought leader”. I didn’t reply, because I wasn’t interested in the job, but I fantasized replying, “Thank you for your interest. Although my credentials as a thought leader are impeccable, I must warn you that I am not that qualified as a female. I can’t walk in heels, I have no clothing sense, and I am not particularly decorative. What aspects of being female are important to this job?”
One thing I definitely think is counterproductive is having women jumping on men to change things that really don’t matter, like whether they use curse words, or tell off-color jokes. How can we be taken seriously if we are such fragile creatures that we will whither away at a curse word? How can people enjoy working with us if they have to worry about a joke being taken the wrong way. That being said, of course I understand real harrassment.
But being too much on the lookout again reinforces women as being hard to work with and focused on irrelevant stuff. What I think will help is having women be successful. But how can we make that happen? When you see coworkers with impediments such as insecurity, be nurturing. Make it safe to ask questions. Encourage women to take risks. Be intellectually curious. Try to learn as much about as many things as you can. Write a book, publish papers.
Since I’m running out of time, I should think of a way of closing. I’ll end with my anecdote that helps illustrate why it’s a good idea to know what problem you’re solving before you try to solve it. When my son was 3 he ran up to me crying, holding up his hand saying, “My hand! My hand!”. I kissed it a few times and said, “What’s the matter, honey? Did you hurt it?” He said, “No. I got pee on it.”