IEEE

Oral-History:Marlene Hazle

SHARE |

From GHN

Revision as of 18:15, 23 May 2012 by Jglock (Talk | contribs)
Jump to: navigation, search

Contents

About Marlene Hazle

Marlene Hazle grew up in Toledo, Ohio and matriculated at Cornell University, where she worked with an early computer system there. After graduating with a degree in government, she moved to Boston to work with the RAND Corporation on their SAGE system. She worked to train new in the SAGE system, and then moved to MITRE to help construct the basics of the first automated air traffic controller system. At MITRE she was also involved in constructing tactical warfare systems for the Air Force and developing internal-use software and specifications.

In this interview, Hazle discusses the development of distributed computer systems, the benefits and problems of working for government contractors, and the benefits of the widespread sharing of codes.

About the Interview

Marlene Hazle: An Interview Conducted by Janet Abbate for the IEEE History Center, 11 February 2001.

Interview #601 for the IEEE History Center, The Institute of Electrical and Electronic Engineers, Inc.


Copyright Statement

This manuscript is being made available for research purposes only. All literary rights in the manuscript, including the right to publish, are reserved to the IEEE History Center. No part of the manuscript may be quoted for publication without the written permission of the Director of IEEE History Center.

Request for permission to quote for publication should be addressed to the IEEE History Center Oral History Program, 39 Union Street, New Brunswick, NJ 08901-8538 USA. It should include identification of the specific passages to be quoted, anticipated use of the passages, and identification of the user.

It is recommended that this oral history be cited as follows:

Marlene Hazle, an oral history conducted in 2001 by Janet Abbate, IEEE History Center, New Brunswick, NJ, USA.


Interview

INTERVIEW: Marlene Hazle
INTERVIEWER: Janet Abbate
DATE: 11 February 2001
PLACE: Hazle's home in Lexington, MA


Family and Educational Background

Janet Abbate:

This is an interview with Marlene Hazle on February 11th, 2001.

To start, can you tell me when you were born and where you grew up?

Marlene Hazle:

I was born in 1934 in Toledo, Ohio, where I lived until I went to college. I went to a public high school, the Thomas A. De Vilbiss High School, and decided that I wanted to go East to a large, coeducational college with high academic standards, and at that time, Cornell University was the only place that fit those requirements. So I went to Cornell. I started out actually thinking I was going to be a physics major, and took physics and a lot of mathematics and sciences like chemistry, but along the way decided that I didn’t want to pursue physics, and ultimately ended up with a degree in government. That was one of the strong departments at Cornell at that time, and I think it still is, and I became interested in that.

But along the way, as I said, I had taken a lot of math; and the way that I got started in computing was through a professor who taught numerical analysis, and said that if anyone in the class wanted to learn how to use a computer to do the fairly laborious homework that was involved in numerical analysis, he would be happy to show us how to do that. And so a few of us in the class took him up on it, and went to learn how to use an IBM computer that was in one of the engineering buildings at Cornell. I actually don’t remember that I used it very much to do the homework, but I thought it was interesting. And this man was well-connected in that very early computing community, and so he passed out the names of the students who had taken him up on learning how to run the—it was the Card-Programmed Computer. And as a result, we received invitations to be interviewed, and it was as a result of that that I was interviewed by the Rand Corporation, which was pretty much at the beginning of its work on the SAGE system, and went to work for the Rand Corporation on the SAGE system in 1956.

Janet Abbate:

Let me back up a little. What did your parents do for a living?

Marlene Hazle:

My father was a mechanical engineer, and my mother was a teacher for a short time before she was married, and after that she did not work. So I was exposed to¬—certainly the notion that I was going to go to college and learn how to do something; have a career.

Janet Abbate:

That was—they really—that was understood?

Marlene Hazle:

That was understood.

Janet Abbate:

Did you have brothers and sisters?

Marlene Hazle:

Two brothers. Neither one of them, though—pursued an engineering degree, which of course would have made my father very happy; so I was—I’m sort of his engineer.

Janet Abbate:

Interesting.

Marlene Hazle:

My older brother is an orthodontist, and my younger brother owns his own plastics manufacturing plant—business. So I was the one that sort of did the scientific thing in our family.

Janet Abbate:

How early did you get interested in science?

Marlene Hazle:

Oh, I guess I was always interested in science; I took all the science that I could take in my high school. I can’t say that I was really focused then; I always liked just about every subject that I took. So, it was—I hate to say this, but it was easy for me.

Janet Abbate:

What’s wrong with that?

Marlene Hazle:

It was easy. I think that had something to do with . . . And I liked the preciseness of it. I liked the fact that a mathematical problem was either correctly done or not correctly done; and the same in the physics and the chemistry: You’d understood the equations and how they worked, or you didn’t. Well, I think that that’s what pushed me towards the more scientific interests, as opposed to the humanities.

Janet Abbate:

Did you already know when you went to Cornell that you wanted to do something in science?

Marlene Hazle:

Well, as I said, I went thinking that I was going to be a physics major. And I enjoyed the physics; I did well; but I guess I honestly thought that I didn’t want to commit myself to going through a Ph.D. program. And it seemed to me that in order to be anything in the field of physics, you really had to do that.

Actually, would you believe that my first advisor at Cornell was Hans Bethe?

Janet Abbate:

No! So that was a pretty high standard to live up to.

Marlene Hazle:

Right! It was a high standard! [laughs.] Actually, I only saw him once!

But I guess I always had it in my mind that I was going to do something in the mathematical or scientific area.

Janet Abbate:

You used that first computer in a class at Cornell. Was that enjoyable? I don’t know how well you remember that.

Looking for Work

Marlene Hazle:

Actually, as I say, I don’t remember how much I did. I thought it was interesting, and understandable. I actually tried to get a job—this was in my junior year—I tried to get a job that summer in Cleveland, where my family lived by this time, having moved from Toledo; but I couldn’t find anybody who cared that I knew what a computer was at the time, so I didn’t! But I certainly, once I began interviewing for jobs in my senior year—with Rand, with IBM—I was convinced that that’s what I wanted to do. It was new, and interesting, and it just seemed appealing. It also paid very well . . . surprisingly enough. Computer—people who knew anything about computers were very scarce in those days, as you can well imagine; so the salary was quite impressive.

Janet Abbate:

Compared to other things you could get with a math degree?

Marlene Hazle:

Well, I didn’t actually—I mean, I’d graduated with no math degree, but I’d graduated with a degree in government. But, compared to what men in the engineering schools were getting, it was a good salary. It was a better salary.

Janet Abbate:

Really? So it was better than engineering!

Marlene Hazle:

Yes.

Janet Abbate:

Great! So what were the job interviews like?

Marlene Hazle:

The people who interviewed me for the Rand Corporation described what the SAGE System was, because that’s what they were hiring for. . . . They were basically selling me on taking the job, because they needed people, and I at least knew what a computer was, and that was a big step upwards for them. And I actually had decided to come to Boston with three friends, and so it was located where I wanted to be—which was a major draw.

Janet Abbate:

Just for interest you wanted to come to Boston?

Marlene Hazle:

These three friends and I decided we wanted to live in a big city where none of our families lived—where none of us had grown up, really—and start out on our careers together. And that’s what we did. And we picked Boston because it was, and it still is, an interesting city, and none of us had come from Boston—or New England, as a matter of fact. We were from mostly the Midwest; three of us from the Midwest and one from New York City.

The interview with IBM was interesting. We were flown to New York City. After an initial interview on campus we were flown to New York City, to some—the IBM headquarters someplace in the city—and entertained royally. But . . . . I was disappointed. Actually one of the other friends (of the three) also was in the same class, had done the same thing I did, and was also interviewing for a computer job. So the two of us were flown down there, entertained royally; but we were unhappy to learn that the only kind of job they were wanting to offer us was what they called the System Service Girl, which was, at that time, wiring plugboards that controlled the operations of the computers, such as the Card-Programmed computer.

Janet Abbate:

So you would go . . .

Marlene Hazle:

And you would go change the program by rewiring these boards for the customers when they wanted to do something different, or when they had a problem. But I guess we both thought that we wanted to something more challenging than that, and so both of us actually started to—chose to come with Rand, and to work in the Boston area.

Janet Abbate:

So, at IBM, if you had been men, they would have offered you a different position?

Marlene Hazle:

Well that was our assumption! I can’t be certain that that’s true.

Janet Abbate:

Did any men that you knew interview there?

Marlene Hazle:

Oh no, we were the only two people being interviewed, and we were—not part of a group in any sense. But the message was, “This is what women do here.” Well, that may have just been that office or that group of people who happened to be interviewing; I can’t assert that it was a more widespread attitude! [laughs.] But that was our experience.

Starting at RAND

Janet Abbate:

You started your first job at Rand—did you say when that was?

Marlene Hazle:

It was in the Fall of 1956. We went to the programming class; and that, of course, was what you had to do initially, because they had this new SAGE computer, which was different from any other computer, certainly different from the one that we had worked on at Cornell. And then I actually went into training immediately. The job of Rand—one of the jobs of the Rand Corporation was to hire and train people to go to the SAGE centers—air defense centers—when they were opened around the country, and do the maintenance of the software: maintain the software, fix it, do the upgrades, whatever was required.

So they had to develop a training program here at the MIT Lincoln Labs. And it just so happened that as a part of our programming class—or at the end of our programming class, I should say—each person, or each couple of people—basically each person—was given one of the SAGE programs that had already been written; and the task was to understand it by reading the code, and then document it so that other people could understand it without sitting down and going through the laborious job of reading and understanding the code—because in order to do that, you really had to understand the application, which in the case of SAGE was very complicated and not familiar to many people. And I happened to be given the code for the operating system—or the executive; that was before the term “operating system” was in use—for the SAGE system, to understand and to document. Well of course, it made a lot of sense for people who were going out to maintain the SAGE software in the field to have some understanding of how all the different pieces of software fit together and are controlled by the executive. And so I developed a course in how that system worked and taught it to the subsequent trainees—hirees—who came in to Lincoln Labs and then subsequently went out to the SAGE centers to help maintain the software.

Janet Abbate:

Was that your idea, to develop the course?

Marlene Hazle:

Uh . . . no; it was not my idea. There was actually a person who had sort of started to do that, and I took it over from him, because he went off to one of the SAGE centers when they started opening them up, and I preferred to stay in the Boston area, having made this big commitment to come to Boston and live. So I was the one that took that over and developed it, and developed another course that had to do with how you start the machine up in the first place, and how you recover if the system goes down—which of course it did frequently in those days! [laughs]

Janet Abbate:

That’s interesting. All of the things you mention—the documenting, and error recovery, and so forth—all those things we know are necessary, but often it seems like after the fact they decide they need someone to do those things. Were those new endeavors, in the sense that no one had really thought, “Well, how do you recover when the computer crashes?” in a systematic way?

Marlene Hazle:

Oh, well I think that the people who were designing the SAGE software—and the hardware, because there were things built into the hardware to facilitate recovery—they had given it a great deal of thought. SAGE was the first real-time, large computer system, so many of those things were new, I’m sure, to the people who were sitting around and specifying the requirements for that software. And those requirements were very thoroughly documented. We looked back at the requirements for the SAGE system: it’s amazing, the level of detail at which the requirements were documented; and the software was written from those very detailed specifications. And when someone like me wanted to understand the executive, I just had to go back to the specifications for it—and then, of course, push a little harder to find out why the specifications were what they were.

I think that what was interesting was that somebody decided—realized that for the people to do the maintenance out in the field—and there were to be thirty-two SAGE sites, so we were talking about many people to do that—a description of the operation of the software in the SAGE system had to be in a more readable form than the specifications were. And also, just one level more detailed than the specifications were. And that’s how—why—that was one of the tasks of the Rand Corporation: to have the trainees document the software initially, and then make it available to the people who came after them.

Janet Abbate:

Were you doing training as long as you were at Rand?

Creating the Air Traffic Control System

Marlene Hazle:

Yes, actually I was. In the fall of 1957, Rand decided to transfer the bulk of its operation back to Santa Monica, which is the headquarters of the Rand Corporation. And so I went to California, with a number of other people, and continued the training there— although I was doing training in other things as well as what I had started out doing (the executive system)—just training in the general operations of the overall SAGE system. Then—it was while I was still working there, but after the System Development Corporation had split off from Rand to take on the ongoing responsibility of providing maintenance people for the SAGE system—I was actually lent back to Mitre—what had become Mitre, because it was split off from the Lincoln Lab—to work on the first air traffic control system prototype, which was implemented on the Whirlwind computer, down in the Barta Building there on the MIT campus. And then I subsequently decided to come back to Boston, and to Mitre, where I continued to work: a little bit on that first prototype, and then was heavily involved in the specification and oversight of the implementation and the testing of the second prototype of today’s air traffic control system—which was implemented on one of the SAGE computers.

Janet Abbate:

That was implemented on the—IBM was making those computers? I forget exactly what they called that.

Marlene Hazle:

The IBM AN/FSQ-7. Just one of those catchy names! [laughs.]

Janet Abbate:

So did you come back specifically because you wanted to work on that air traffic problem?

Marlene Hazle:

I came back because I wanted to come back to Boston, and I also very much liked working on the air traffic control. Again, it was something new and different. And working on a significant prototype system was very exciting, and gave one a wonderful opportunity to understand how to—understand the application for which you were writing software. I actually went to air traffic controller school in Oklahoma City to learn what air traffic controllers did! —and used that information in helping to write the specifications for our prototype system. And then when the prototype was completed, and we were ready to test the system, the FAA sent, I think it was thirteen working air traffic controllers to participate in our test and evaluation of this system for a year. They came and lived in Boston and participated in the test and evaluation of that system for a full year.

So we had a marvelous opportunity to really understand what the application was, and how real air traffic controllers could or could not use what we had built for them. Of course we did lots of revision based on their experience trying to use the software that had been developed for them. Well, that’s the purpose of a prototype, of course, to put something before somebody and try to use it to do whatever the job is, and then understand where its shortcomings are, and how it can be improved to better support the individuals.

Janet Abbate:

So your job was quite broad, in the sense that you needed to thoroughly understand the application, figure out what needed to be done, and how you could translate that into . . .

Marlene Hazle:

Into some software, some displays. Right.

Janet Abbate:

Right. And then, in addition, doing the actual coding or implementation or figuring out what hardware was needed—I mean, were you personally involved in every stage of that?

Marlene Hazle:

I was not personally involved in the implementation. Actually, the implementation for this second big prototype air traffic control system was done by the System Development Corporation, and Mitre was responsible for the specification. SDC was basically a subcontractor, a contractor providing the software. Then we would run the system; we would test it to make sure that it was doing what we thought it should be doing—that it had met the specifications—and then the air traffic controllers would try to use the system. And we actually—we actually had airplanes flying around in the sky who were part of our test, so that we knew where the airplanes were, and we were tracking the airplanes using the radar inputs that were coming into the system, and then the controllers were trying to keep track of the planes and to perform their control function.

Yes, so it was broad. It was very broad, from the understanding of the requirements in the first place, then writing what they’re trying to envision—you know, how those requirements could be met through displays and interactions—and then writing those specifications, and then testing the software, as I said before—testing the software to make sure that it did what you specified, and then testing it at another level, to make sure that in fact helped the controllers do their job.

Janet Abbate:

That must have been exciting. When was the first time you took an airplane that was using your system?

Marlene Hazle:

Oh! Well! It was quite a while later, because the prototype was completed in—we finished the work on that prototype in 1964; and then the Mitre people who had been involved in this prototype were required by the FAA to go to Washington, to be close to the FAA headquarters, to specify and oversee the acquisition of the real air traffic control system. And so I did not participate in that, because I chose to stay here in the Boston area, and it took some time for those specifications and the implementation and the testing to get done—of course, there were new computers to be built. So I can’t honestly remember when that air traffic control system first became operational. It was a number of years later.

Janet Abbate:

What were they doing until then?

Marlene Hazle:

Oh! Well, they were talking on the radio, and . . .

Janet Abbate:

And using radar?

Marlene Hazle:

They did have radar screens, but they were just—blips being displayed. There was no automated ability to follow a series of blips and say, “Ah! that’s an airplane!” And, “That’s American Flight 369 that’s going from Boston to New York!” And, “He’s where he’s supposed to be” or “He’s not where he’s supposed to be!” And it’s that kind of support that the software of the air traffic control system gave the controller that he didn’t have before—you know, as well as many other things. So they were on the radio and they were just looking at these sort of basic, raw radar returns; and they were counting on people to fly from one navigational aid to the next and in the air lanes that they were supposed to be in.

Janet Abbate:

What was your next big project after that?

Working With the Air Force

Marlene Hazle:

After that, I was involved in the development of yet another prototype system; actually I spent many years involved in the development of prototypes at Mitre. And this was a prototype system developed to demonstrate—really, it was more of a demonstration system, I would have to say—how computers could be used to support command—commanders, Air Force commanders who were trying to understand a war situation, in particular a tactical situation. So, we were using a very sophisticated, new computer at Mitre called the Stretch, and we were trying to put together a system fairly quickly, because as I said it was really a demonstration system, and so in the process of doing that we were encouraged to use what was then a new kind of programming language, a list-processing language.

Janet Abbate:

Was this LISP?

Marlene Hazle:

It was a derivative of LISP. It was called TREET, and I don’t even remember what TREET stands for! But anyway, we did—and it was developed at Mitre; this language and system was developed at Mitre—so we did develop a demonstration system, and many high-ranking officers in the Air Force were brought in to participate in—I think it was a two-day exposure to the kinds of things that computers could do for them, if they had good displays of situations and good displays of what their resources might be for dealing with the situation. And out of that we developed yet another prototype system, which was specifically oriented to tactical air mission planning for the Air Force. Now, this was the time when the Vietnam War was going on, and there was of course a great deal of interest in providing some computer support to the mission planning function. And we developed something that was called the Tactical Data System Testbed. It was actually a very early implementation of a distributed processing system, where we had a central computer, which was a rather small IBM Process Control Computer, to which was attached a PDP-8; and attached to the PDP-8—there were eventually two PDP-8s—were workstations—what were called “intelligent workstations.” At each one of these workstations there was a display; there was a tape recorder, a printer, and another little gadget, which enabled people who couldn’t type on the keyboard to sort of pull computer messages—messages that they wanted to send to the system, to request display, or to print, or to do some of the other things that they were able to do—This little gadget allowed them to take pretyped messages from a little tape cartridge and put them up on the screen, so all they had to do was type in the variable values! [laughs.]

Janet Abbate:

Sort of a menu.

Marlene Hazle:

Right.

And we developed—in the process of developing this system, we developed a very simple high-level language that allowed us to—oh, generate displays very easily, so that we could get the kind of quick reaction time that you want when you’re doing prototype development. You want to be able to take a look at something and say, “Oh! Well, I wish I’d done that some other way!” and then make the change as rapidly as possible.

So, we again brought in people—a limited number of Air Force people—to help us make an assessment of the usefulness of this tactical air—tactical data system—to support the mission planning function. And then some of us who were involved in that prototype development ultimately went on to participate in the specification of a real Air Force system that would support tactical data planning. So it was transferring the knowledge gained from the prototype into the next generation of system that would in fact be procured by the Air Force and put into use, eventually.

Janet Abbate:

And so that was actually used at some point?

Developing Large Systems

Marlene Hazle:

Oh yes, it was. Of course, these things took a long time! I think—I’m just sort of guessing now, as I look back—the sort of normal cycle was: You started writing serious specifications for a system—now, this was after you’d done some prototyping and done other investigations of what the requirements really were—from the time you started writing those specifications until the time that you might have sort of the first version of that system that you were specifying: a minimum of five years. And often substantially longer than that. And then all those systems went through various phases: you would start out with a very limited subset of the total capability that you had specified, and get that up and running—in the field, ideally; and then add a little more functionality on. And those systems lived for ten, fifteen, twenty years.

Janet Abbate:

What was the biggest challenge, from your perspective, in creating these kinds of systems?

Marlene Hazle:

Well, to me the biggest, the most interesting part was in fact the understanding of the application, and the envisioning of how that job might be supported with the—now, we’re talking, all along here, interactive systems: displays, and messages, light guns, light pens—the precursors of mice! [laughs.] So to me, the challenge was to understand the job, and then to envision what kind of software might be useful; and then to have the opportunity to see and evaluate it in a prototype situation. Unfortunately, as time went on, there was less and less money to support these significant prototype developments, so you had less and less opportunity to envision a solution, have it implemented, and then get the feedback of actual user—prototypical use. There were many demands for different systems, and limited resources—both financial and personnel—and so the prototyping tended to be minimized in system developments.

Janet Abbate:

Do you think that’s a change in general, over time? Not just at Mitre?

Marlene Hazle:

Yes. I think over time the people working in the computer field—at least some people—got better at making that leap from the application to useful specifications and useful software. But I think all of us who had the opportunity to participate in those early prototyping efforts lamented the lack of emphasis on that over time, and felt that systems were being specified without—and procured, and put in the field—without adequate understanding—without adequate testing of the efficacy of what had been specified and implemented.

Now, one of the things that I was involved with, which was sort of a way of addressing some of these problems, was the development of guidelines for the user-system interface of these display-based systems, so that for each new system being specified and implemented people weren’t starting over with a blank page and saying, “Well, I wonder what the display should look like,” or how the information should be arranged, or how the interactions between the system and the users of the system should take place. So guidelines—we actually undertook the development of a substantial set of guidelines for people to follow when they’re designing displays for computer systems. And of course, people were starting to write books about how to do systems engineering, which is what this big process was generally called. So, some of the knowledge was being captured in books and papers and things like that.

But, as I say, I think most of us who had ever participated in those prototypes felt that it was unfortunate that—in our case, the Air Force, which was a major customer here in Boston—needed to go out and specify systems without having had the benefits of the prototyping.

Janet Abbate:

And the input from the users.

Marlene Hazle:

And the input from the users.

Janet Abbate:

Who just got stuck with whatever got built.

So that would be, about, in the ‘60s that you were starting to try and codify a bit the user interface?

Marlene Hazle:

Oh, no, not as early as the ‘60s! I don’t think so. I mean, people were sort of writing down things, but I think the writing down of the guidelines for the user-system interface was, oh, somewhere I would say in the late ‘70s, early ‘80s. That’s sort of my best recollection of it. It took a long time. . . .The people in the computing field . . . didn’t do a lot of writing! [laughs.]

Janet Abbate:

I guess you’d have to build a few systems before it occurred to you that that was an issue.

Marlene Hazle:

Yes.

The person who did the major work in this—the development of user-system interface guidelines—was not only an engineer, but actually had a doctorate in clinical (I think it was clinical) psychology. He wanted a great deal of certainty before setting out to write down guidelines for other people to follow—a lot of corroboration—so I think there was a hesitance on his part, for sure, and on the part of other people, to write down rules unless they really had a lot of data and consensus to support the rules that were getting written down. So I think the natural cautiousness of people with a scientific background—to have the data to support what they said.

Janet Abbate:

Well, it’s a tough area.

Marlene Hazle:

Because I remember many people would say, “Well, give us some guidelines.” And it was difficult to get them written down, for a while.

Janet Abbate:

Did you ever find yourself wishing you had some new piece of hardware or something to put in the system? Like, “Gee, I wish we had a light pen,” or something?

Marlene Hazle:

I don’t remember . . . I don’t actually remember, myself, looking for hardware to solve the problems, so I can’t say I’ve actually had that experience. I was much more interested in finding the software to support our own work, actually . . .

Developing Internal-Use Software

Janet Abbate:

You were developing software tools for your own internal use?

Marlene Hazle:

Right. After having worked for a number of years on a couple of these large systems that I’ve talked about—the air traffic control, and the tactical planning, and then some support for the military airlifting plan—Military Airlift Command Planning—I, and Mitre, became of aware of the emergence of some tools that might be useful to us in doing our own work, and might be useful to the contractors who were implementing the systems that, basically, we had specified. And so we started looking into things like program design languages and specification languages; and we set about doing some evaluations of some of the things like program design languages and specification languages—and the supporting software that went with them, that had become available—and decided that, well, nobody else had quite the same job that we did, and therefore, in particular the specification languages and capability that was available—the most well-known one, and probably the furthest developed one—was really not adequate to support the kind of specification writing we did.

And so we undertook to develop some tools in-house to support the work of Mitre and the Air Force, and this was a research and development effort. And we did develop a system that helped us put together the written specifications for systems, utilizing—actually utilizing text that had been generalized from previous systems; text that would describe things like parts of the user-system interface. And you would then tailor these blocks of text to meet the specific requirements of whatever specifications you were writing, but it was sort of that head start, so you’re not starting with a blank page again— because there’s a great deal of similarity, particularly in the military systems that we were involved with, going back to the SAGE air defense system.

There were a number of different versions of that system that were developed subsequent to the original SAGE system. And some work on systems that had a lot of similar functionality, although not necessarily air defense systems. For example, many systems were involved in tracking airplanes. So the idea was to use this information—the knowledge, the experience that had been developed in earlier systems—and make it available to the people who were specifying the follow-on systems.

So that was one kind of support we provided.

Janet Abbate:

Was there a name for the specification language?

Marlene Hazle:

Well, we didn’t develop, actually, a specification language, it was just—that part of the system, which we just called the “specification development system,” was simply a mechanism for storing and indexing these components, or reusable elements of specifications. Now, I’m talking a time prior to word-processing systems. This is before the PC! [laughs.] We also developed the capability . . . Well, actually, not quite prior to word processing, because Mitre was involved in word processing on mainframe computers very early on in the game. We developed a series of tools that allowed one to—would you believe—spell check? And index, and perform lots of sort of simple functions that any word processing system would do today, but at that time there was no automated support for that.

And then, as another piece of software that we developed to support the work of Mitre, we developed a system—you know, another component of this specification system—that helped us to manage the comments and modifications that were made to specifications, because when you’re writing specifications for a big, complex system, you go through a number of drafts of these specifications. You send out the draft and the users comment on it, and the hardware people comment, and anyone who has a stake or a part of the system has the opportunity to comment on the specifications. So you get these reams of comments back, which need to be reviewed, evaluated, and disposed of: either incorporated into the specifications or not incorporated into the specifications, and if you don’t incorporate them, you have to be able to answer why you didn’t incorporate them. So we developed some software to help us manage that important function.

And that’s about the same time that we developed these user guidelines, as a matter of fact. So that was a period of having the opportunity to be our own users, and to go through the iterations based on our own experience with the use of the system, because we actually went out to help the other people within the Mitre/ESD [Electronic Systems Division] community use this software that we had prepared to help them. Because it was actually a fairly complicated system, which just took some getting used to, in how to use it, and it involved lots of people .

Janet Abbate:

Would that actually generate—could you translate specs into code?

Marlene Hazle:

Oh no. The product here was a written-type specification, which would then be used as the basis for procurement. Contractors would read those specifications and propose hardware and software that would meet those specifications, and then there would be a source selection, and the Air Force would pick one of the bidding contractors, and then they would go forward and—still, again, using the specifications—actually implement a system that they had proposed to meet those [specifications].

MITRE Systems

Janet Abbate:

How was the SAGE system, and the systems at Mitre that were kind of inherited from that: How was that different from what other people were doing in computing, as far as you could tell?

Marlene Hazle:

Well, of course, as I . . . That’s where I started out, so I just assumed that it was standard. But over time I began to realize, when I started going out of my own situation and meeting other people in the computing world— (There weren’t a lot of conferences then. At the very beginning there weren’t any that were specifically dedicated to computing; for a while they were mostly engineering things, I think, where the computing people got together.) —It took me a long time to realize how unique the SAGE system was, in that it was so big and so complicated; that it was a real-time system—we were actually trying to process radar as it was coming in, and respond to the actions of operators in real time, so that when they punched a button, something happened; it was a very early, sophisticated display system; early interactive mechanism; and in lieu of the sort of typewriter keyboard that is standard today, there were banks of buttons to be pushed, which of course had to be programmed for their functionality; and, in lieu of the mouse as a pointing device, there was what we called a “light gun,” which actually looked like a gun, and was pointed at the display. So it took me quite some time, actually, to realize that most people were doing FORTRAN on mainframe computers, and using punch cards and tapes, and that they were not engaged in this sort of—the real-time interactive computing systems that we were.

Janet Abbate:

Did you get to program them interactively, as opposed to batch processing?

Marlene Hazle:

Well, I never had programming as my chief job. I was just always either in training early on, and then in the requirements specification, and that sort of thing. Well, I did do some programming, but it was always a small part of one of these prototype efforts. And that was interactive, yes. Yes.

Janet Abbate:

So that was unusual.

Marlene Hazle:

Yes.

Janet Abbate:

Was that a fun place to work?

Marlene Hazle:

Yes! It was fun because we were always doing something new. At least, for many years we were always doing something new. And I think by and large, the people who were working on these systems were people who wanted to do a fine engineering job, and who wanted to do something useful for the customer—whoever the customer was, whether it was the FAA, or whether it was some Air Force officer who was trying to do planning, or trying to do air defense in the case of the SAGE system—so there was there was lots of personal interaction with the customer, particularly in the early days, when we were involved in so many of the prototypical developments. I always said I’ve never done the same thing twice, which was certainly true in the beginning and almost true to the end! So that kind of job is always challenging.

I’m a little stumped to think of anything [else] . . . I’m inclined to say: You know, part of the fun was that the people are so bright—you know, who were doing this—and they are, I think. Mitre’s always described as sort of quasi-academic, and what you get is a lot of the real bright people who aren’t trying to make a fortune, but trying to do something useful, something good for society, or the country.

Janet Abbate:

Sounds great!

Marlene Hazle:

And, so it . . . Although . . . There certainly were pressures. There was lots of pressure, because people were so notoriously bad at judging how long it was going to take to do anything. You know, over and over you would—people would make commitments to do something by a certain time, but in fact they were unrealistic commitments, but—as is the case with human beings, I guess—whether it’s unrealistic or not, you don’t want to fail to do what you committed to. So there were those kinds of pressures.

Women in Computing

Janet Abbate:

Would you be working all night?

Marlene Hazle:

Sometimes. Not too often, but . . .

Janet Abbate:

How did you balance work and family?

Marlene Hazle:

Well, I’m not married, and I’ve never been married, so I didn’t have to make that balance. My life was mine to choose to do with. Well, I certainly had an outside life: I was not working like one of the dot-com kids. No way! But every so often we would get into a situation where we’d end up working all night, or most of the night, usually for two or three days, something like that—but not on an ongoing basis.

Janet Abbate:

Were there a lot of other women in your group?

Marlene Hazle:

At the very beginning, my programming class—my SAGE programming class—I would say it was about a quarter women. And I would say that, say, over the first . . . . Well, the time I was at Rand and SDC, there continued to be a high number of women. You know, these are people of course who are coming from all sorts of different backgrounds, to be trained in programming and to go off and find a new career. I mean, there were teachers, there were . . . Among the women, I would say . . . Well, I’m not sure . . . Several people I remember were teachers, and the rest I can’t . . . Well, and a number of them were just college graduates like me . . .

But then, at Mitre, there were few women. There were very few women in the beginning.

I would say that probably 70 to 75 percent of the people at Mitre are engineers—have engineering degrees. That’s probably still true, and that certainly was, I’m sure, true–even truer—then. And so, there are so few women in the engineering field that you would sort of expect few women— And actually, the other women that were at Mitre early (like Judy [Clapp] and me), none of them were engineers; none of us were engineers. We had all, you know, come from various academic backgrounds. Over time, of course, some women engineers did come, and then there began to be computer science graduates, after a number of years.

I never personally felt—I never really sensed discrimination or bias, in terms of not being able to do something I wanted to do, or being, you know, in any sort of personal relationship with somebody, a situation where I felt that there was bias. But underlying, there was a bias against promotion.

Janet Abbate:

Were salaries comparable, as far as you could tell?

Marlene Hazle:

As far as I know, they were. Among the members of the technical staff, I don’t think there was significant bias, but there was a bias against promotion for a long time.

I don’t know how unique Mitre would be in this respect. But—and this probably was somewhat unique—at Mitre there was no—for many years there was no management development program. There was no mentoring; there was just—It was: “You can do it on your own.” And in order to—in order to get someplace, like to be a project leader or to be promoted, you really had to go after it. Well now, that’s true to a certain extent. It wasn’t entirely true. But there was no— As I say, there was no development program, for anybody. And I always felt that it was a mistake for Mitre, as a company, to not have that sort of thing. I always felt that there was a lot of wasted potential there, because it was a do-it-on-your-own . . . .

Janet Abbate:

Were they subject to affirmative action policies, since they did so much government contracting?

Marlene Hazle:

Yes.

Janet Abbate:

And did that change things?

Marlene Hazle:

Yes, it did: somewhat. It made them . . .

Janet Abbate:

In the seventies, maybe?

Marlene Hazle:

I can’t tell you . . . but it probably was the seventies. I’m very poor at remembering when things happened.

It made them much more conscious about hiring, interviewing, and actually sort of aggressively looking for female candidates—because it was really female oriented at that point, rather than minority—and that was the first push. And it made them take a look at salary discrimination, and I’m sure they did some adjustment of salaries, when affirmative action was a very strong influence for any government contractor.

I think that was sort of the beginning of opening their eyes to the reality of the male domination!

Janet Abbate:

Do you think that’s partly because they did so much military work—the male focus?

Marlene Hazle:

Well, I think that it probably affected the extent to which potential candidates were interested in Mitre. In terms of whether it would affect the people at Mitre—whether that would affect them—I wouldn’t think so, actually. I think it was just sort of “this is the way it’s always been.”

Janet Abbate:

Engineering culture.

Marlene Hazle:

Yes.

Janet Abbate:

Do you think computing jobs have become more open to, or welcoming to, women?

Marlene Hazle:

I’m not sure I can answer that question, just because I don’t think I have a broad enough experience. I spent my career really in one place, and so I didn’t see—directly, through direct experience—different situations; although I did get to know people at a number of large corporations, people who worked for contractors on some of our systems. And I’ve felt that as time went by, there actually was a decrease in the number of women that I would see at conferences.

And I never was too sure why that was. I had a theory that in the beginning, nobody—everybody was a beginner. There was no hierarchy; the new companies didn’t have any ingrained hierarchy. But as time went by, those hierarchies began to develop, and it was the men who were at the top. And also, as more established companies began to make computing a part of their work, again, they already had an established, male hierarchy. And I always felt that that probably had something to do with the apparent decrease in the percentage of women that I encountered.

And then I must say in the last—say, the last ten years that I worked, I began to see a lot more women again. These now, of course, were the younger women coming into the work force with degrees in computer science, almost all of them, or engineering—a few in engineering—and in general, quite—at least the ones that I had much contact with—quite strong personalities, and strong technically. So that was pleasing to me, to see some young women coming in.

Janet Abbate:

You finished at Mitre in ‘83?

Marlene Hazle:

‘93.

Janet Abbate:

‘93, sorry. So that would be the ‘80s and early ‘90s that you’re talking about.

Do you have any advice for young women contemplating going into computing?

Marlene Hazle:

[pause.] Well, I guess it would probably be my advice to any young person getting into any new career, and that would be: to make sure that you’re interested in what you’re doing, and that you understand where what you’re doing fits in with the bigger picture of computing, and understand what a progression—a professional progression, from wherever you are at that point—what the progression is. Be looking to see where you’re going. And I would strongly recommend—and have recommended to my nieces, one of whom is an aeronautical engineer, but has had a lot of involvement with computing, and the other one who is working in the computing business for an accounting firm—to find a female mentor: because there’s still few enough females in the field that you could benefit from having the advice and support of somebody who’s a little ahead of you.

One of the things I say in retrospect on this subject is that strangely enough, the few women at Mitre never sort of banded together in any real way to form any kind of support group or mentoring function. And I’m sorry that that didn’t happen for those of us who were very early, just in terms of being with each other and talking to each other; and I’m sorry that we didn’t try to help the people who came after us, in a sort of deliberate way. I think we would have benefited, and I think the younger people would have benefited. I mean, I did have some relationships with younger women, but it was a very—just because we happened to like each other, and be working together—it wasn’t any premeditated sort of relationship.

The Benefits of Sharing Code

Marlene Hazle:

And I would throw in this comment that I may want to erase, and that is that, in the beginning, strangely enough there was a lot of competition among people in the computing field at Mitre—and I had a certain amount of sense of this outside of Mitre, too—that everybody wanted to sort of be right —be, you know, be the best, and so there was competition and a non-supportive attitude for a very long time. Among groups of people, more than on an individual basis, but among groups.

Janet Abbate:

That’s sort of reminiscent of what I hear about how male computer hacker culture was very competitive. Do you have a sense that it was sort of an engineering/computing kind of culture? Or some other reason?

Marlene Hazle:

It was male. It was, you know, I would say . . . It was a sort of a male thing, and there was lots of sniping, and . . . . Where everybody was so new—I mean, ideally you would want to share anything that you learned with other people; but that’s not what happened.

Janet Abbate:

Judy [Clapp] had mentioned something about people being very possessive about their code—and I guess you weren’t maybe coding as much, but . . .

Marlene Hazle:

Right: She was much more involved in that than I was.

Janet Abbate:

Someone had written this book called Egoless Programming —

Marlene Hazle:

Right.

Janet Abbate:

—to kind of counteract that prevailing thing; which surprised me, that it seems like there was kind of this more territorial feeling at the time.

Marlene Hazle:

Well, right; and your saying that about Judy [reminds me . . .] Right. I think that people were very possessive about the code they had written, because it was sort of a laying-out of their mind and their intelligence on paper that other people could see. And they didn’t necessarily want anybody else looking at it! And God forbid that somebody else should come in and change it, or tell them how to change it!

And that’s an interesting thing too, because—I told you that the very first thing that I did, and that the other people who came to be trained by Rand did, was to read the code that had already been written. Now, I will tell you that was the best education for how to program—was to read somebody else’s code—good code, ideally—and to see how it’s done. And so many people were really denied that opportunity because of this “it’s my code” kind of stuff. I mean, I always felt that it should be a part of anyone’s education and training to sit down and read good code.

Janet Abbate:

That’s so interesting, because I interviewed a woman, Elsie Shutt, who ran her own programming business . . .

Marlene Hazle:

Right . . .

Janet Abbate:

It’s called Computations, Inc., and it was women with children working part-time from home. But anyway, what she said was that it was their standard practice that they would swap code and read each other’s code—mainly for error-checking, you know, before using expensive computer time—but that it was really beneficial, because they all learned tricks and stuff from each other, from reading each other’s code all the time. And I remember that she really emphasized that, and it sounds like that really could have been a useful thing.

Marlene Hazle:

Absolutely!

Janet Abbate:

Did that change at some point at Mitre, that atmosphere?

Marlene Hazle:

The antagonism? I would say that it subsided as the field matured. At another level of our work there was a corporate effort that was initiated maybe ten years before I retired. Somebody up high realized that people had not been talking about their system engineering and management mistakes. The mistakes were being made over, and over, and over again! And somebody said, “Well, you know, the reason that we’re making the mistakes over and over again is that nobody wants to talk about the mistakes. So they always document the wonderful things we did here, but they don’t document the failures, and an analysis of, ‘Well, why did something fail? What could we have done differently? And how did we deal with the failure?’”

So there was a big push to document lessons learned, and to share information—both lessons learned in a positive sense and lessons learned in a negative sense. And that—I think Judy is still involved in something, a sort of a corporate capturing of the lessons learned—because people just realized that you save yourself a lot of trouble, time and money, if you share information about, particularly, the problems, and how they were caused, and how they were solved. So I think that there has been a recognition of the value of that. And there’s also—I mean, there have been some organizational changes at Mitre that were just being initiated at the time I retired, I think—to support sharing of expertise, and people, resources, more than in the past. But it was a long time in coming.

But I remember very vividly the antagonism, the sniping, that just seemed so unnecessary.

The Joys of Computing

Janet Abbate:

Well, to look at the other side: what have you found to be the most satisfying aspects of your work with computers?

Marlene Hazle:

Well, I would say: the support for human activity that makes things easier and more efficient. Deep down somewhere inside me I have a sort of drive for efficiency and straightforwardness, so I think that’s always been the pleasing—the thing that pleased me about the work that I was involved in.

Janet Abbate:

It was kind of aesthetically satisfying?

Marlene Hazle:

Yes! Yes. There are a lot of people in the computer field who are very—particularly the computer scientists—they like the elegant solution, they’re into elegance. I’m into functionality and simplicity—well, elegance can often be simplicity too, but . . .

I guess I’m much more—I’ve always said—I think (and I’ve read this, too): I look at a computer as a tool. It’s—I’m not a game player; I’ve never played games on computers, and I don’t look at the computer as sort of a fancy toy. To me, it’s a tool. And I think a lot of women look at computers that way, as opposed to—a lot more men, I think are interested in the “toyness” of it: something to play with; that wonderfully complex, infinitely malleable thing that computing is—as opposed to looking at it as a way to do something.

I know somebody wrote some books about that a few years ago. There was a woman—I would say maybe ten years ago, fifteen years ago—there was a woman in academics who was trying to explain the difference between women in computer science and men in computer science.

Janet Abbate:

I don’t know if I know that . . .

Marlene Hazle:

I can’t tell you anything more about it . . . I think, actually, there was an article in the New York Times, and I remember—I can see the picture of the woman, but not clearly enough to know who she was.

Janet Abbate:

I’ve got a sort of title that won’t come to the front of my mind that might be that book, but I’m not remembering it. And I think that—I’ve certainly heard other people say similar things.

And in your case, you were actually able to build systems that eventually got used by real people with real needs. The “toolness” was built in.

Marlene Hazle:

[laughs.] It was pleasing when we were making tools for ourselves, too, because then we got to use them!