TechStuff Remembers Grace Hopper

of Rear admiral in the United States Navy. However, she did not die while in service. So I want to make clear that this episode does not dismiss the countless people who have given their lives in service to the United States Military, but rather to remember someone who made enormous contributions to technology, particularly while a member of the US military. And we also associate her with a story that usually gets told incorrectly. In fact, I've been guilty

of telling it incorrectly myself in the past. The person I'm talking about is, of course, Grace Hopper, who sometimes gets the affectionate nickname Grandma Cooebol. Note I said cobol Cobol, not cobold, as in the Nasty Little critters in Dungeons and Dragons. Grace Hopper was born in nineteen oh six in New York City to a well off family. Her father, a Yale graduate, owned an insurance company, and she grew up not wanting for anything really. She attended private schools

and she matriculated at Vassar. She focused her studies on science, specifically in physics and in mathematics. She graduated Vassar in nineteen twenty eight, and then she followed in her father's footsteps. She enrolled in Yale for graduate studies. In nineteen thirty, she earned her master's degree in mathematics. She continued her graduate studies and pursued a PhD. Around the same time, she also accepted a job to teach mathematics at her alma mater, Vasser. Her mentor during this time was a

fella named Howard Ingstrom. He was just four years older than Hopper herself. So just a quick word about Ingstrom. The story is not really about him, but we need to at least get an idea of who he was, and it is a bit important. While he was a mathematician, his actual degrees were in chemical engineering, and then he would go on to be one of the co creators of the UNIVAC computer. Now, keep in mind at the time we're at right now, with Grace Hopper's timeline, we're

in the nineteen thirties. The concept of an electro mcanennic computer is really just starting to become coherent. Around this time. There were people working on it, but these were very very early days with the idea of an electro mechanical computer. Now, there had been earlier examples of purely mechanical computational machines like Babbage's difference engine, Babbage's analytical engine. Those were at least concepts that were computational engines that used mechanical components

to calculate things. Considering where Happer's career would take her, having a mentor who was an early contributor to computer science and design is meaningful and they would have other encounters in the future. But back to Hopper herself. She had the opportunity to study with the German American mathematician Richard Courrant, whose work in the field of applied mathematics was renowned, and she earned her PhD from Yale in nineteen thirty four, and she earned it in mathematics and

mathematical physics. So then we skip ahead a few years. On December seventh, nineteen forty one, Japanese forces bombed the US base at Pearl Harbor in Hawaii. The United States, which had resisted entering into the conflict we would call World War two, found itself drawn into the conflict, and Hopper wanted to be part of the war effort, and

she applied to become enlisted in the Navy. However, the Navy rejected her, partly because Grace Hopper was of a petite build, diminutive in size some would describe her, and also because at age thirty four at this time, the Navy considered her to be too old to join the effort. Nevertheless, she persisted, and Grace Hopper would end up joining the

US Naval Women's Reserve in nineteen forty three. So while she wasn't an active member of the Navy, she was an active member of the Navy Reserves, and she received an assignment to work on the Bureau of Ship's computation project at Harvard University. There, she worked with a Harvard

graduate named Howard Aiken. When Aiken was a graduate student at Harvard, he lamented that his doctoral thesis relied upon complex calculations that were tedious to perform by hand, and so, like many innovators before him, he decided the best option was to find people he could work with to invent a device that could do those kinds of calculations automatically.

So he was able to work with a relatively young company called International Business Machines better known today as just IBM, and together they would be able to bring his wish into reality. So Iiken would kind of act as the director or at least the guide officer, like he's guided the development of this machine to perform these calculations automatically. The result was a device called the Automatic Sequence Controlled Calculator ASCC, but later and better known as just the

Harvard Mark one. So Hopper joined the Bureau of Ships right around the time that the Mark one had been shipped to Harvard from IBM. So the Mark one didn't look anything like a modern computer does. It was an electro mechanical device. It had moving parts like there were shafts that would turn and gears that would turn, and

all sorts of stuff like it was. It was a real machine, and it also had electrical parts right there were electrical switches, there were various relays, and there were miles and miles of cables, something like hundred miles it was. You know, I've seen different estimates, but yes, hundreds of miles of cables connecting all these different components. The whole thing collectively weighed more than four tons. If you were to look at the deal and get a tape measure out,

it was about fifty one feet long. It was eight feet tall and then two feet deep, right, so like you might want to think, it was like fifty one feet wide, eight feet tall, and then two feet deep something like that. That's a definitely not a desktop computer unless you have a truly ginormous desk. This computer could store a few dozen numbers, and you could actually then

use those numbers for multiple calculations. So that was useful, right, right, You didn't have to insert the same number every time you wanted to do a different calculation. You could actually store a number in this device. In fact, more than seventy numbers at a time, and I think it was something like it was more than twenty decimal points or decimal digits that you could store in this per number.

The length of time it would take to generate an answer would depend upon the nature of the calculation, so simple operations like addition or subtraction would take around a second for this machine to perform. More complex mathematical functions could require a minute or more. But it was all automatic. At least it was once you input all the data, which you would do by using switches. Lots and lots

of switches. The Mark one had more than a thousand switches in various switch banks, so you had to use all these in order to input the data you wanted to then operate upon. Now to tell the computer what you wanted done to the input, remember inputs just one thing. You also have to instruct the computer what operation to perform. Well.

To do that, you would use a program, and in those days, a program consisted of you know, punch cards in some cases, but in the case of the Mark one, it was essentially paper tape that had holes punched into it, and you would feed the tape through the computer and those holes would allow for specific connections within the computer, and in turn that served as instructions for the computer. It would perform a specific operation depending upon which contacts

allowed to go through. So programming a computer meant physically punching holes in cards or in tape, and Grace Hopper would be one of the first three people to do this. So she was one of the first three computer programmers for the Harvard Mark one computer. Now she wasn't just a programmer, She never was just a programmer. She was also an instructor. Remember, she had come from a background of teaching. She taught at Vassar before she joined the Navy Reserves. So it became her job to write a

user manual for the Mark one. The story goes that I Can initially didn't want to give Hopper very much responsibility because he mistakenly believed that due to the fact she was a woman, she would be limited in that capacity. And then Hopper proved him wrong, and so then he said, all right, you know what, you know what you're doing, write the instruction manual for the Mark one. Now, as you might imagine, the complexity of the Mark one required

a pretty long instruction manual. From what I understand, the manual had more than five hundred and fifty pages of instructions. Yikes. Now, keep in mind again it's an electro mechanical system, really complicated. Literally lots of moving parts, which means there's lots of potential points of failure. So it makes sense that any comprehensive manual would need to be long. But can you imagine being put in charge of writing a comprehensive instruction

manual for such a complicated machine. She did it, and she would work on various projects that would contribute either directly or indirectly to the war effort. So the lab she worked in, the calculations they were performing using this computer, they were for all sorts of things, including things like numbers, tables for stuff like rocket trajectories or specific types of artillery, so that soldiers in the field would have a reference and they would know what settings to use when they

were targeting enemies. Right like you know where the enemy is, or you're told where the enemy is in relation to where you are, you have to actually make the calculations of how to position, say a piece of artillery, how to position the gun barrel, so that when you fire, you're actually going to hit that target and not something else, Right Like, you don't want to end up hitting a civilian building. You're aiming at a military unit or or

something like that. You need these tables so that you understand, based upon the equipment you're using and the firepower behind it, that you are actually going to hit the thing you're aiming at. That's the kind of calculations that Hopper's team was working on, and some of the work that was done with the Harvard Mark one would also end up being used by the Manhattan Project that was the top secret mission to develop the atomic bomb. They also would

depend upon calculations made by the Mark one. Now, at the conclusion of World War Two, Hopper would choose to stay with Harvard. She was given the opportunity to go back to Vassar, but she decided she would stick with Harvard and work on computer systems. We're going to take a quick break, but when we come back, we'll talk more about the contributions Grace Hopper made to computing. Okay,

we're back. So, as I mentioned before the break, after World War Two, Grace Hopper decided that she was going to stay at Harvard, at least briefly. She would stay there so she was an intrinsic part of the team that developed the Mark two, the successor of the Mark one, so a more complicated and sophisticated computer. She also worked on the design of the Mark III, which obviously followed the Mark two, and at this point Harvard was still receiving support from the US Navy, which had a vested

interest in having access to sophisticated computational machines. Around this time, we get to a story that has been mangled quite

a bit in the past. In fact, even Yale's own page about Grace appears to have a mistake in it, because the Yale page says that this incident that we're about to talk about happened in nineteen forty five, but according to other sources that talk about the actual log book that recounts this event, it happened on a specific day two years later, on September ninth, nineteen forty seven.

So again, this is a story that has been told and retold many times, and a lot of those retellings are inaccurate, probably not intentionally inaccurate, but it just has happened, And like I said, in the past, I've repeated some of those inaccuracy. So I'm going to try to be as careful as I can to recount what actually happened. Keeping in mind that this story has been told so many times that it's very possible that there is no one hundred percent accurate version of the story. But here's

how it goes. So the Mark two team at Harvard were running into problems. The Mark two computer was producing error after error, and it was doing so consistently. So there was something wrong and it didn't matter how many times they ran the calculation through. They were getting the same errors at the other side, which meant something somewhere in the computer system was hanky. I mean, maybe a

connection was loose, maybe a wire had broken. Like there was no way to know unless you cracked the sucker open and really took a look. So that's what Hopper's team did. They began to look through the Mark two computer to see what was going on, where was this problem actually occurring. Then they found it. A moth had been trapped inside the Mark two computer and it got zapped in a relay and thus it was blocking a signal. Right, the relay could not relay signals because there was a

a dead moth in the way. So they removed the x insect and they actually taped it into the log book they were keeping where they were. You know, they would write down whenever they had errors you know, they would log them so that they could work the errors out. So they taped the dead moth into the logbook, and they included the amusing note of quote first actual case

of bug being found end quote. So the shorthand for this story is that Grace Hopper invented the term bug when it comes to a problem that's in a computer program or even in a computer system. But that's not entirely accurate. So the term bug meaning something is going wrong inside a mechanical or electrical system, something is creating errors. That pre dates the nineteen forty seven incident. That was a term that was already in use in various engineering circles.

And that makes sense, like it has to have already existed or else that log entry doesn't make sense. Right, The log entry was first actual case of a bug being found, So that implies that the term bug was already being used to describe errors. It's just that in

this particular case the bug was a literal bug. So a lot of historical accounts now say Grace Hopper was the first to apply the term bug to a problem within a computer system, not just an electrical system or mechanical system, but a computer system itself, and effectively she coined the phrase computer bug. That's possibly true. It's at least more accurate than just saying she coined the phrase bug,

because that's not true. Also, we should mention that it's really her team, Like to this day, there's a little bit of co of who first made the joke about it being a bug. She also gets the credit for the term debugging, and initially that makes sense because her team literally had to debug a computer system by removing the moth from the relay. However, the term debugging was

already being used in other circles as well. In fact, it appears in a nineteen forty four letter that Jay Robert Oppenheimer wrote to Ernest Lawrence, for example, that happened three years before the Great Moth incident in the mark to computer, So debugging was already a thing too. You could argue maybe she was the first to use debugging in relation to a computer system. And maybe she was,

or at least her team was. Now, as far as I can tell, Grace Hopper never claimed that she or her team coined the term's computer bug or debugging, that this was something of a narrative that develop around her but she was not responsible for it. She never made those claims. Now, I think it might be because the actual story of a real bug causing problems in a computer is so amusing that folks want it to be the origin story for the term itself, right, because it's

such a good story. Wouldn't that be awesome if that's why we say computer bug? And it's still a great story, right, it's fantastic, and it really did happen to Grace Hopper's team, So not taking anything away from that, it's just that it's not where we get those phrases. It's just an amusing example of it being literal in this sense. Now, we're not done with Grace Hopper, not by a long shot, because her contributions went far beyond a close encounter of

the moth kind. That's just the amusing story folks like to tell that. It relates to Grace Hopper's contributions in computer science. So Hopper faced professional challenges. Despite her impressive work to both the war effort during World War Two and to computer programming in general, the Navy denied her a regular commission, so she decided to leave active service.

She would remain a reservist until nineteen sixty six. More on that when we get there, But she was no longer an active service member of the US Navy at this point. Not long after she stepped down from active service with the Navy, she actually chose to leave Harvard as well because she had been denied tenure, and it also became obvious to her that she was not going to get promoted there despite all the work she was doing.

That she was just hitting a glass ceiling super hard and there was no real chance of her getting beyond it. So rather than bang her head and or accept her fate and not get any further in her career, she decided in nineteen forty nine, she left Harvard University and she went to join a computer company, a fledgling computer company called the Eckert Mauchlee Computer Corporation or EMCC. The following year, a typewriter company called Remington Rand would acquire EMCC.

And this is where we're going to get into a very complicated history of a very important computer system. All right. So EMCC, or Eckert Mauchley Computer Corporation, had taken its name from its two founders, Jay Presper Eckert and John Mouchley. Now, these two had built another early computer. This one was called the Electronic Numerical Integrator and Computer or ENIAC, at the University of Pennsylvania. ENIAC was also an important computer, or at least it was to be a important computer

during World War Two. The actual development and construction of ENIAC took quite a long time, but ENIAC is one of those famous early electrical computers or electronic computers, i should say. Now. Following that achievement, the two men found themselves at a fundamental disagreement with the University of Pennsylvania over the matter of patent rights, and this led the pair to leave the University of Pennsylvania and then go

and found their own company. Thus, EMCC and EMCC designed a new computer called the Binary Automatic Computer or BINAK, which is essentially just a footnote in history. There's very little about BINAK out there. It's possible that the machine was literally never used for anything. However, it did arguably leave the Bureau of the Census to then seek out EMCC to create a new computer design, which the company

would refer to as UNIVAC. And Grace Hopper would join EMCC around this time and would be important in the work and development of UNIVAC, specifically in programming for the UNIVAC. But in the fall of nineteen forty nine, the primary financial backer for EMCC died in a plane crash. So once their financier essentially had died, EMCC was kind of in financial trouble. They weren't sure whether they could last long enough to build out the UNIVAC. That's when Remington

Rand came along and acquired EMCC. Rimmington Rand was, really, like I said, a typewriter company. They made office typewriters, but they saw the opportunity to get into the computer world by purchasing EMCC. However, there's always a bigger fish, so things are about to get way more complicated. I'll explain what I mean by that after we come back from this quick break. All right, we're up to nineteen

fifty one. Grace Hopper has been part of EMCC and then Remington Rand, which purchased EMCC for a few years, and that's when the company completed work on the first UNIVAC computer. It would eventually be known as the UNIVAC one, but at the time they just called it the Univac. Kind of like how in World War One. People referred to it as the Great War because if they had started calling it World War One while they were fighting it, that would have been, you know, very pessimistic. Calling Univac

the UNIVAC one would have been very optimistic. So they built the UNIVAC, and unlike other computers, this was not a computer that was just meant to be a one and done like. They made multiple UNIVAC one computer systems. The first one went to the United States Census Bureau because that's kind of where all this work got started,

but the company made other ones for other clients, including CBS. Famously, CBS would use the UNIVAC one to predict the outcome of the nineteen fifty two presidential election, but when they got the results, they didn't trust it because the Univac had said that Eisenhower was going to win handily, that it was going to be a landslide victory for Eisenhower, but the conventional wisdom at the time was that this

was going to be a much closer presidential race. However, it would ultimately turn out that the UNIVAC prediction was right, like, not exactly right. They didn't get to the very you know, the very voter or anything like that. But it was way more correct than the guests that it was going to be a tight race. And that ended up really elevating univacs prestige because once CBS said, Okay, well it turns out our computer predicted a landslide. We just didn't

believe it. But it turns out the computer was right. That's amazing press for your computer system. Right. So, Rimington Rand slash EMCC would continue to develop the UNIVAC platform, and then Rimington Rand made another acquisition. It acquired a competing computer company called Engineering Research Associates or ER. So Rand acquired ER in nineteen fifty two. Now, one of era's founders was Hopper's old mathematics mentor, Howard Ingstrom. Rand

effectively ran EMCC and ERA as separate entities. Right, so, while both companies were under the corporate umbrella of Remington Rand, they still were kind of competing against each other. They were making different computer designs. That, however, would change in nineteen fifty five. I told you this gets complicated. So in nineteen fifty five, Rand merged with another company called Sperry spe r r Y, and then the new company

was known as Sperry Rand. Now, at this point EMCC and ERA divisions would be merged together to form a new division called Sperry UNIVAC. One of the folks at ERA really didn't like that because it meant that ERA had to drop a lot of the work it had been pursuing. This guy was known as Seymour Cray. He left the Sperry Rand Company and decided to make his own computer company. It's called Cray Computers. That's another story. It might mean something to you if you know the

names of old supercomputer systems. Yeah, we'll have to tell the story of Cray Computers in another episode. So anyway, at this point, the head of the division that was Sperry UNIVAC was a guy named Leslie Groves. Now, Groves had been an officer with the US Army Corps of Engineers. He was actually in charge of the construction of the

Pentagon back in the nineteen forties. He also served as the military's director for the Manhattan Project, and he went on to oversee the early days of what would become the United States nuclear weapons program. But he had also reached an impasse with the US government. He was told outright he was never going to get the appointment he wanted, and in the late nineteen forties he decided to leave

military service. He received the rank of lieutenant general just before he retired, and then he entered civilian life and he joined Sperry and became a vice president of that company. So I mentioned that again because this episode goes out a Memorial Day and Sperry Rand this computer company had a really healthy representation of former Armed Forces personnel working

as engineers. There. In fact, this was not unusual across computer companies of the time that the engineers who were at the forefront of the United States development of computers were people who had previously worked in that respect for

the Armed Forces. Now, as I said, Hopper was involved with the work on UNIVAC primarily from a programming side, and her team would create something that would push computer science in general and computer programming in particular into a new era and to make it way more accessible for new programmers. And this was the first compiler. All right. So I've talked in the past about how machines do

not understand languages the way we do. So when you get down to the basic operational level, machines don't understand human language right, Like, you can have a conversation with chat GPT and the appearance is that the machine understands what you're saying, But this is actually after several layers of abstraction. When you strip it all away, machines understand machine code. Typically we're talking about binary code, so that means at the basic level of computing, you're dealing with

zeros and ones. Machines can process this information efficiently, but it's not easy for humans to work with that. Now, early programming languages were really mathematical codes, and we're very, very close to giving to machine code itself. Grace Hopper had this dream of making a system where humans could write instructions for computers in a more easily accessible language,

something that was closer to the language like English. Now a lot of are contemporaries dismiss this as just wishful thinking, like, yes, it would be nice if you could program in a language closer to English, but it would also be nice if I had my own private plane and could fly anywhere I wanted to do, any time of day. It's just a pipe dream. How can you bridge the gap between what humans find easy to do when it comes to communication, and what machines are capable of, and Hopper's

team got to work on creating a solution. Now. That solution came in two important parts. One is programming language. Now, programming language creates a level of abstraction from machine code to make programming less of an arduous task, but not all programming languages are alike. A low level programming language is really just a few degrees separation from machine code itself.

It's still pretty challenging to learn. It is a slog to really get a firm grasp on how to program, which in turn means you can be really limited in what you can do right that you have the potential to do much more interesting programming, but you have a limitation because the language itself is so difficult to work with. A high level programming language has more layers of abstraction, so it makes it easier for humans to grasp and to work with, and you're also less likely to introduce

bugs like programming bugs. Obviously that still happens, but it's slightly easier to avoid. But to make a programming language useful, you have to have a way to translate those instructions from the more human friendly programming language to the useful machine code that the computer system actually relies upon. To that end, Hopper's team built the first compiler, and the compiler does pretty much what I just described. It takes code that's written in a programming language and compiles that

into machine readable instructions. The early programming languages, like I said, we're essentially mathematical code like Fourtran is an example very close to mathematical code. Hopper's team developed a programming language that began to approach something similar to English, and this was a huge step. It was a non trivial development in the world of computer science. Hopper's team named the language Flowmatic. Unlike predecessor's like Fortrand, Flowmatic's use of English

commands remove some of the barriers to programming. It was still a challenge to learn. It wasn't like the easiest thing in the world to work with, but the learning curve became way way less harsh when you started to introduce this more English structure type of programming language. Flomatic would essentially become the inspiration for a future programming language called Common Business Oriented Language or COBOL, which we mentioned at the top of the show. Or COBOL if you prefer.

And Hopper really liked Cobol, and she promoted it and worked on compilers for the Cobol language, and that helped make Cobol a widely used language in academia, in the private sector, in the military, and her work really made huge results. People would later attribute her promotion of Cobol as a major factor in the language's success, because by the nineteen seventies it would become the most commonly used programming language out there, and people say Grace Hopper's contributions

were a large part of that. Now, as I mentioned earlier in this episode, Hopper was in the Navy reserves until nineteen sixty six, so she had left active service, but she had remained in the reserves for two decades until she was required to retire due to age restrictions, and she would reference her retirement from the Navy as the saddest day of her life. Her rank at the time was commander. And then seven months later, the Navy came back to her and said, Grace, could you help

us out again? And after retiring as a reservist, Grace Hopper would rejoin the Navy as an active service member.

Her responsibilities primarily included streamlining the Navy's approach to computers and to find a way to manage the multiple computer systems and computer languages the Navy was reliant upon, because by this time, the Navy had invested in multiple computer systems, working with lots of different people who took different approaches, and it just became complicated and difficult to coordinate projects because you had different systems working with very different computer languages,

they were not inherently compatible. So Grace Hopper was there to try and make sense of it all and to make it more efficient and streamlined. She would spend nearly two more decades as an active service member of the Navy, and she got the nickname Amazing Grace. She also continued to work for UNIVAC until nineteen seventy two. Throughout her career, she earned countless distinctions, way too many for me to

list in this episode. At the age of seventy nine, she finally retired from the Navy for realsi's no takebacks. By that point, she had attained the rank of Rear admiral, and at that point she was also the oldest active service member in US military, which is pretty pretty phenomenal. As I said, she got numerous awards and distinctions, both while she was alive and posthumously. She has received awards from US presidents, from various universities, from various professional associations.

An incredibly important person in the world of computer science. She lived until the ripe old age of eighty five. She passed away on New Year's Day nineteen ninety two and was buried an Arlington cemetery with full military honors. So truly a phenomenal figure in the history of computer science and a testament to what you can achieve when you have ambition, and you have skill and talent, and

you have dedication to your craft. A phenomenal life's story, and so I am glad to have dedicated an episode to Grace Hopper. I hope all of you out there are well. I hope those of you who are observing Memorial Day have a peaceful day of reflection. And I will talk to you again really soon. Tech Stuff is an iHeartRadio production. For more podcasts from iHeartRadio, visit the iHeartRadio app, Apple Podcasts, or wherever you listen to your favorite shows.