§1. Course logistics and presentation schedule [00:00]
The proposed schedule: we'll start the week of October 19th on the presentations. Remember you get ten overheads, ten minutes, we'll have a few minutes for a discussion of whatever you present, and people can ask questions. Hopefully we'll get about three done an hour. This hasn't really been a problem in the past, and we may try to go between nine and eleven to get six done a day, and be done within about a week or six days.
Professor Slocum asked me to give you his — he called it a hyper apology — he thought the class started at ten. We're trying to reschedule, see if he might come in one of the last couple of days. Simone has listed the schedule for the rest of the semester. We should finish on October 15th with the lectures — the 24 lectures, 12 from him and 12 from me. I'll be lecturing Monday and Friday of next week, then I'll be doing the thirteenth and fourteenth, and we should finish on the fourteenth.
§2. Scientists vs. engineers: who solves which problems [01:10]
We were talking about engineering involving trade-offs and decisions, and I'm starting to talk about how you make decisions, which really gets into how you think. I talked to Dave Ragone this past week. Anybody working on batteries? Nobody. There's something called a Ragone plot, that David Ragone — Dave's about 85 now. He used to teach in the Department, he was a graduate of this department and he went on to do several things. His comment was: the scientist solves problems that can be solved, and the engineer solves problems that must be solved.
A scientist, if there's a problem they want to solve and they feel they have the capability of solving, will simplify it until they can solve it. An engineer on the other hand doesn't have the luxury of simplifying the problem. You've got to build the bridge all the way across the river; you can't build it halfway. If there's a problem with going all the way, you've got to figure out how to complete the job. That's why a number of professors in engineering schools are merciless on problem sets. If you don't get what they define as the right answer you get zero — no partial credit. Because to an engineer, doing the job even partially incorrect is a failure. You've got to get it a hundred percent right. The scientist, well, "in the right ballpark." But the scientists will write it down with differential equations and everything else, and so they feel very confident with their final solution. That runs into problems sometimes.
The example I was going to tell you is Brian Josephson. Anyone know who Brian Josephson is? You ever heard of the Josephson effect in superconductivity? Nobody's ever heard of it. Well, Brian Josephson was a graduate student at I think Cambridge rather than Oxford — anyway, a British university. He was a brand new graduate student, like 21 or 22 years old, and he went in to see his professor. His professor was sort of busy and said, here, go solve this problem. This was a problem on electron flow that people had written down the equations before and solved many times. It was just sort of an exercise to keep him busy and for the professor to see how well he could do on a graduate-school-type physics problem. Well, Brian was naive, Brian didn't know. He wrote down the terms in the differential equation for the forces on the electron as it's flowing through a material, and he wasn't smart enough to know that one of the terms is usually just thrown out and set to zero. You've done those solutions, right? You look at the differential equation and you say, oh, this term is insignificant, we'll drop it.
Well, Brian was actually a good enough mathematician that he solved the problem with the term that everyone else simplified out. He actually solved the harder problem, and he discovered the Josephson effect, which is basically electrons tunneling through materials. That was 1962, and before he was 30 years old he had won the Nobel Prize. How about that. When I met him in the early 70s he was sort of a basket case. He was in his early 30s and everyone was just fawning all over this brilliant person, and he just didn't know how to take it. He was actually a pretty humble guy, didn't have much personality, didn't know how to take all these people fawning all over him. He's a nice person. But there's an example: in general the scientists all knew you throw out this term; another person comes along, doesn't know that, is too stupid to know — and he wins the Nobel Prize for being too stupid.
There's also the quote from Frank Whittle, the British engineer who developed the first turbine engine. He did this in the 1930s and early 40s, and there were predictions by other scientists that this is impossible. Remember Lord Kelvin's quote, something about powered flight being impossible. Well Kelvin was wrong. And it turns out people were telling [Whittle] that you couldn't build the turbine engine, it wouldn't work. His quote was: it's a good thing I was stupid, I was too stupid to know it wouldn't work. Because he made it work. He was an engineer. So one of the differences is, the scientist will simplify the problem to where it can be solved; an engineer has to figure a way around the problem.
Another thing came out recently. MIT Spectrum came out last week, and here's the quote from Rafael Reif in his introduction: "MIT, solving problems together for the benefit of mankind." That "for the benefit of mankind" is sort of an engineering definition. The scientists are not solving problems for the benefit of mankind; they're solving problems to increase human knowledge, which they say will benefit through somehow in the future in some unspecified, vague way. "MIT focuses on problem solving in service to the world. We stop worrying about boundaries between disciplines and focus instead on what it takes to solve the problem — making new tools, seeking new perspectives, inventing new solutions." They solve the problem that must be solved, not the problem that can be solved. He also talks about interdisciplinary creativity, complex challenges — remember, the school of engineering defines engineering as complex, ambiguous, uncertain problems — and bonds of mutual respect. So this ties together some of what I've been talking to you about. I didn't even prime him to say these things.
§3. Kahneman: System 1 and System 2 [08:00]
Understanding how we make decisions means understanding how we think. I started to talk about Kahneman's book — this psychologist, I guess PhD. As someone pointed out, the difference between psychiatrists and psychologists is one's an MD, can prescribe drugs. Kahneman won the Nobel Prize for trying to figure out how people think, and he came up with this system one and system two.
[Tom puts up a summary slide.] People have written synopses of Kahneman's book. This is a 30-minute summary, because it's a pretty dense book to read. It'll take you a while — save it for Thanksgiving vacation. His book will put you asleep.
In chapter one, he basically says there are two main processes in the brain. They're not really different parts of the brain, but as people have studied things, they can divide the system into a system one and a system two. System one is intuitive and automatic. When you see stuff in bold and capital letters, you don't have to make a mental effort to read "Thinking Fast and Slow." If you see a stop sign you know what that means. You develop all kinds of habits in your own life — when you brush your teeth, whether you eat breakfast, how you get dressed. Things people develop become what's normal for them, and that simplifies a lot of decisions in your everyday life.
System two on the other hand requires skill, attention, and effort. There's a fundamental assumption in Kahneman's book that the brain is lazy. We like to find heuristics or habits to allow us not to have to make a decision about everything that happens in our life. If we had to analyze everything that occurs, we wouldn't have time to do anything else. So out of pure survival, system two is not invoked except when system one can't deal with the problem. If you saw a ball of fire form in the corner over there for no reason, we'd all have to start analyzing — what happened, where did it come from, we've never seen that before, it's not within our normal experience.
Chapter 3: self-control and deliberate thought are both forms of mental work for system two. Some people can experience a mental state of flow, a state of joy considered an optimal experience. Has anyone ever had that? I was just talking to a guy — had breakfast with him, had dinner with him at the beginning of the week, today's Friday night — and I asked him why he worked so hard. He says, because of the rush when he solves the problem. I remember when I was doing a lot of research and had a dozen graduate students, a student would come in and we'd work on something, they'd show me the results, we'd talk about it — we're doing system two thinking. And all of a sudden the light goes off and you realize you just solved the problem, you now have a framework for the world and you could publish a paper on this. There's a real joy. I've felt the flow, he's felt the flow, and the reason he works hard is he likes to get that rush. Maybe it's better than heroin, I don't know, I've never taken heroin. But it's their thing.
He goes through associative activation, where one idea in the brain triggers other ideas — the priming effect. If someone tells you it's going to rain, and all of a sudden you see water coming down the stairs, the rain caused the flood, because you're already thinking about rain. You're primed to explain things in terms of something you just saw recently. A lot of the politics in the world that we disagree on is due to priming. We see the American flag and we think patriotic — this is good, we share our wealth with the world. When we were donating food to different parts of the world, we'd ship the rice or the wheat in bags that had the American flag all over them. We'd get to Ethiopia, and those people saw that as the worst evil in the world, because the American flag to them is primed as people who have come in and wiped out parts of their country. So we're primed to think of the American flag as something good, and other people are primed to think of it as something bad.
The whole ISIS thing: those people have been primed to think about certain things in certain ways, and we think about things in different ways, and we just don't even understand how they make their decisions. We think, oh, they're destroying these ancient antiquities. Well, for them, on the ladder of abstraction or the Maslow scale, it's more important to be religiously pure than to preserve some ancient art. Whereas we think ancient art is very valuable and we'll trade it for millions of dollars — to them it's not important. He goes through lots of these things, and the reasoning won the Nobel Prize because his work is so general — it gets into how we think. This will be on Stellar.
§4. Heuristics in practice: the Johns Hopkins pad eye [15:25]
There are heuristics we develop, and I talked about the one I used when I was looking at faculty resumes: if it took them too long to graduate from graduate school, they're not going to be very productive as a faculty member. I'll show you another heuristic that I used, because I haven't been giving you very many of the engineering stories. A number of years ago I was asked to evaluate why a pad eye had failed. I'll show you what a pad eye is in a second.
Johns Hopkins University has an Applied Physics Laboratory, which is sort of like MIT Lincoln Laboratory. The Applied Physics Laboratory is mostly funded by the US Navy. A guy got hit in the head by a piece of metal while he was aboard ship out in the Pacific — a research ship being operated by Johns Hopkins. It was towing on a cable some little object that we weren't allowed to know what it was because the project was classified. We do know there was a submarine involved. It was a sonar test. They were running a sonar experiment. Right there in Annapolis, Maryland at that time was the head sonar laboratory for the US Navy, right across from the Naval Academy. I used to spend a couple of weeks there each summer. I was in the welding building, and across this nice courtyard was the sonar building. On the second floor was all the hush-hush stuff and the computers. You could walk through the first floor — just a bunch of pumps — and you could walk anywhere you wanted. I always thought security was sort of lax there, and I mentioned it. They said, oh no, someone once walked through the wrong door to get up to the second floor, and within two minutes there was a Marine with a gun behind every tree pointing at that building. So they do some classified stuff for how to detect submarines.
[Tom sketches the towed object and pad eye on the board.] So they're towing this object and there's some submarine involved. Up on board ship there was a pad eye. A pad eye is nothing more than something that has a hole in it. This piece of steel was about one inch thick and maybe five or six inches on a side, with a one-inch hole in it. It had a weld that welded it to the bulkhead of the ship, and they had a steel cable going through it. That pad eye — the weld had broken, and it shot across and hit this guy in the head. He wasn't killed but he had some brain injuries, and at the time he was also in prison for other things. He filed suit against the government, I think, or Johns Hopkins. If you looked at this thing, that round circular hole was now elongated in the direction the cable would be pulling.
So, system one thinking: which failed first, the weld or the pad eye hole? The pad eye hole deformed, the weld broke — what came first, the chicken or the egg? The hole. Why? Isaac Newton said something — for every action there's an equal opposite reaction. What's allowing you to bend that hole — before you broke the weld, or after you broke the weld? The pad eye is just a loose trinket on a chain after the weld breaks; nothing is going to elongate a one-inch hole in a one-inch-thick piece of steel after that. It has to be the hole that elongated first. Of the five or six experts that looked at this, they all determined the weld failed first. I'm not kidding. I was the only guy that said, of course the hole failed first. One of my heuristics is: if a piece of metal deforms before the weld breaks, the weld was a good weld. Because the weld doesn't have to hold any more strength than allows the metal to deform. That's one of my heuristics that I've developed for system one thinking. But I'm a welding engineer, what can I say.
Then there's system two. To differentiate these, in system two I had to do some analysis to prove it to these other people. Here's our pad eye up on the ship. They had ten-foot swells in the sea that day in the Pacific. The cable was on the order of two thousand feet long — I knew it exactly at the time, this was years ago. To analyze this: when the ship goes up to the top of the swell, the cable has to stretch, because this object is way down at the bottom. There's a time delay — the mass of the thing they're towing is not going to follow the swells on the surface when it's 2,000 feet below; it's just going to stretch the cable. I wanted to know the load in the cable, because I had to have enough force to deform that hole. That was my hypothesis, and I was going to try to convince these other engineers that the hole came first and not the weld failure.
So ten over two thousand is one two-hundredth — you can do that in your head — and that's equal to the strain. You could do this as force-displacement, but it's easier for me to think stress versus strain. The stress is proportional to the strain if this is all elastic — Young's modulus, 30 million for steel, times 0.005 strain gives 150,000 psi. If that was a one-inch-square cable, you'd be talking about 150,000 pounds of force. 150,000 pounds might bend that little circle, right? In fact I did the calculation, and it would. So that was the analysis — that was work. The example Kahneman likes to use is: you can do a math problem in your head by system one. He uses 13 times 37. Most people will just ignore that and not try to multiply it out, though they will do 2 times 2 automatically because they know the answer. Most of you don't remember the answer to 13 times 37; it takes a mental effort and you usually won't do it unless you're forced to. So those are some of the types of things Kahneman goes through.
§5. Surviving at MIT: career anchors and heuristics for students [23:03]
I've been sort of using this class as a guinea pig. I didn't decide to teach this "What is Engineering" until August. I was going to teach about casting, the module I taught four or five years ago, but finally about halfway through August I decided I'm going to do "What is Engineering." So the reason this is somewhat disjoint is I'd never taught it before and I didn't have a lot of time to prepare. I've been assembling little pieces I've been doing over the years — which is why I gave the leadership thing; I did that 20 years ago, I wrote the article in 2003 on leadership and management education at MIT. You're getting a piece here and a piece there. And it was partly my reading Kahneman this summer that got me thinking, there are some interesting things in common and it would be worthwhile for the students to read it.
A year and a half ago I started writing a book called 50 Years at MIT. I've been here for 47 years and I've got a few more years to finish it — actually seven more years, because I'd like to have it done for my 50th reunion, which is 54 years since I came here. I've written about 70 pages and I'm about halfway through my junior year, so I've got a few more years to go. But I decided to start writing down some of the heuristics that I've developed about how to survive at MIT.
[Tom distributes a handout.] Here it is — the only thing I've ever put a copyright on other than my thesis, which MIT required me to. "Surviving at MIT." Some students have asked me to come and give a talk before; I use this as the theme of a 20-minute talk. You're getting the two-page version with 22 things on it.
The first one: be humble but don't be humiliated. One of the biggest problems for MIT students is you think everybody else around you is so brilliant and you feel like you're not up to snuff, when in fact you're in the top 0.03 percent of the population. You should learn to be humble and recognize that no matter how good you are at something, there's probably someone here who's better. So be humble, but don't be humiliated. Just because someone's better — measure yourself against yourself, not by what others think about you. Have the self-confidence to know that you're a worthwhile individual.
Work hard but recognize your limitations. Don't procrastinate — I've told you that several times. I think ninety-five percent of the pressure at MIT is self-inflicted. Have self-confidence; recognize that MIT is an artificial environment and you're well above average. Here, you're just average. Accept failure, move on. MIT is an elite institution — don't apologize for it, but use it to benefit others. If you're an engineer, serve others. Find your career anchor.
[Tom holds up Schein's Career Anchors workbook.] How many of you have already gotten Career Anchors, anybody? Three of you have a copy. They cost thirteen dollars, so don't take one if you already have one. This is Ed Schein, a professor at the Sloan School. It's a 30-minute quiz on which you will not be graded. You answer 40 questions on things like: "I want to be so good at what I do that others will seek my expert advice — never, seldom, often, or always — one, two, three, or four." Or: "I am most fulfilled in my work when I have been able to integrate the efforts of others towards a common task — one, two, three, or four." Just be honest with yourself. If you want, you could Xerox those three or four pages of the quiz, so someone else could use the book afterwards.
There's a professor at the Sloan School — actually his name is on this edition, the fourth edition — John Van Maanen, who goes around giving this in a half-day or full-day seminar to companies on finding your career anchor. He probably makes forty or fifty thousand dollars a day giving these little seminars. That's why these books cost so much. Ed Schein spent decades thinking about these types of things and came up with eight career anchors. A career anchor is the one thing you will not give up when faced with a choice in your career or your life. It could be lifestyle — some people want to be a park ranger. It could be technical competence — you want to be a scientist or an engineer solving problems. It could be managerial competence — you want to help organize other people to solve problems. For me, believe it or not, my career anchor is autonomy. I don't like having a boss, and I have tenure, so I don't really have a boss. I do have a boss but I can ignore him — it's called tenure. So I'm in the perfect job. I don't have to retire, I enjoy what I'm doing — my career anchor is autonomy.
So there are eight career anchors. Security: a lot of people who work for the government want the security of knowing they'll have a pension, because the government can just keep printing that money. All the rest of us, our savings will go worth nothing, but Congress will keep their pensions at some buying power. So you can find your career anchor. If you're married, have your spouse do it separately, and then figure out what the two of you have. It's important to know what motivates you and those close to you.
[Tom holds up another handout.] Here's another one-page synopsis. Automatic system one, and attentive system two where you have to analyze things. System one's quick, impulsive, intuitive. System two is deliberate, cautious, deferential, effortful — but it's also lazy. There's priming, anchoring, the availability heuristic, confirmation bias. I found Kahneman interesting enough to make it part of this course.
§6. Logical fallacies and a deposition trick [32:41]
There's another thing I've been using for 15 years. I was teaching a freshman seminar on things you needed to know, and I came across this book Fallacies and Pitfalls of Language: The Language Trap by this Canadian, [Morris] Engel, a logician. I've put on Stellar — or will be on Stellar — my synopsis of that. So I wrote my own CliffsNotes for the students, and the date on it is December 2000. I'm just paraphrasing the book. Aristotle basically came up with three hundred and some logical arguments, all but a couple of dozen of which he found were incorrect logically. You can start drawing Venn diagrams of the logic. I'm not handing out the book — you can get one yourself, or if you Google "errors of law" you'll find this document. I was surprised somehow that it's on there.
Equivocation: everything that runs has feet, the river runs, therefore the river has feet. Or, "the miracles of science," just by itself, is equivocation — science doesn't deal with miracles, science deals with facts. If a tree falls in the forest and no one is present to hear it, does it make a sound? That's sort of a philosophical thing. Or I like the Prairie Home Companion version: if a man speaks in the woods and no one hears him, is he still wrong? Prairie Home Companion makes all kinds of jokes about husbands and wives. Sweeping and hasty generalizations, complex questions — you've heard the one about, when did you stop beating your wife, when did you buy your first Cadillac. There's an assumption in a question like that: you bought a Cadillac once, or you're beating your wife. So it's a question you shouldn't try to answer because it's too complex to answer.
There are a number of logical fallacies, and now, particularly when I'm doing my consulting and I'm on the defense and some expert writes a report, I'll often use this reference as a scientific reference on what's wrong with his argument. So it's not just "he's an idiot, of course the hole failed first and not the weld" — you can't write that. You say he's got a hasty or sweeping generalization, chapter 8 of Engel, and I've given a scientific reference for his logical fallacy.
That's the course I've been debating for nine months. Students have been asking me to give a course on my forensic consulting, and I've been advised by several people: don't do it. If you watch all 120 hours of my lectures you'll get most of the secrets, but you're going to have to compile them yourself. If I were to condense them into just a little 12-unit course and put them on the web, every attorney that ever deposes me is going to watch all of those, and they're going to know all my little tricks.
I'll give you one of them right now. An attorney will ask you a question under oath, and after you give an answer explaining it, he'll say, no, that was a yes-no question, I just wanted yes or no. The first time, I give a polite answer: I can't just give a yes-no answer to that question. Yes, I want a yes-no, it was a yes-no question. Sir, I'm under oath and I was sworn to tell the truth and the whole truth — to give anything other than my explanation would not be the whole truth. Some of these guys are really aggressive, and if they want to keep it up — I want a yes or no answer — I'll say: sir, are you soliciting me to lie under oath? Because for an attorney, that's a felony. I've only done that two or three times in my career, but when I start saying "are you soliciting me to lie under oath," boy, the room goes quiet. So there are little tricks you learn over time. You try to be polite, but if the person tries to get aggressive, you still try to be polite — but after a couple of strikes, they're out. One time a guy tried to stiff me on my bill, and I actually used the "solicit me to lie under oath" line, and all of a sudden they came to the settlement table. Just changing it from a civil complaint to a criminal complaint did it.
§7. The scientific method: a church oven fire investigation [38:40]
So, the scientific method. When we make decisions we use a number of tools. Kahneman talks about how we think from a psychologist's point of view. Engel talks about it from logic — pure logic — which you can write down as mathematical formulas, Venn-diagram-type things. And we also know that the scientific method is a proven method of discovery. So that's what I do in my forensics — I'm using the scientific method to go back and analyze a problem. I'm trying to make decisions on how this thing occurred.
Let me give you an example from a couple of nights ago. There was a fire in a church in New York, in the kitchen in the basement. You have this little mom-and-pop shop in Ohio that's selling fifteen-thousand-dollar commercial ranges and ovens for cooking bread and heating things on the stove. Since this is a materials course: the guys who designed these fifteen-thousand-dollar commercial ovens, it's basically just stainless steel sheet metal — and these are sheet metal workers, they're not trained engineers. [Tom sketches the oven.] The oven — this is the back, this is the front with the door — underneath there's a thermomagnetic valve which runs the pilot light. This is sort of an old-style thing, it's 12 years old. They ran this in stainless steel, there was no insulation in this — I don't know why but they didn't have any insulation on the walls, so it's just a sheet of steel on the bottom. And they ran an aluminum tube full of gas for the oven up to this valve, screwed it in.
When the firemen came in, they saw the fire underneath the oven. They'd extinguish it and the flames would go out when they hit it with CO2 or water or whatever, and then as soon as all the steam from the water went away, the fire would come right back. They did it again with their hoses, and ten seconds later the fire comes back. They did this three or four times. They knew there was a gas leak. The only thing in the area of the gas leak with a source of gas was this pilot-light tube made out of aluminum. So what's the problem? This is an oven that operates at 500 degrees Fahrenheit — you've got a piece of stainless steel, and you've got an aluminum tube right beneath it with no insulation in between. This is in service for 12 years.
When you connected the tube to the valve, they used a little brass or steel nut, I don't know which, but it wasn't an aluminum nut. They had a little brass ferrule, a little ring you crimped with the nut to tighten down like a tourniquet on the tube to make a seal. Well, every time you heat that thing up, the aluminum expands more than the brass or the steel, because it has a higher coefficient of thermal expansion. When it cools back down it's now shrunk, and I'm loosening the nut and loosening the crimp on the ferrule every time I cycle this. So let's do an estimate. Twelve years in the church, how many hours has that been in service? Student: [response] On the order of thousands. Let's assume 500 hours a year — say you use it a couple of times a week for five hours, that's 500 hours. How many thermal cycles? I estimated a hundred or so — twice a week. These are just estimates. 500 times 12 gives 6,000 hours of service.
Then I go look up the creep behavior of aluminum — this is system two. System one was just kind of realizing, oh, I don't usually put aluminum tubes inside a hot oven. Not a good idea. That's sort of system one level. System two is to go do the analysis. It doesn't matter whether it's a heat-treated aluminum alloy or a non-heat-treated aluminum — the heat-treated alloys, over a couple of thousand hours, will lose their temper at three or four hundred degrees Fahrenheit, so they no longer have the strength they should have. I don't know the alloy because the whole thing melted in the fire — I have to write a report about something that doesn't exist. Well, I can write a report about something that doesn't exist because it was just a lousy job of material selection. If they'd used a steel tube, which is what they use now, it wouldn't have happened.
You don't have to prove everything in a court of law — well, in a criminal case you do, but I've never done a criminal case. You have to prove what's most probable. We know there was a gas leak there, the firemen saw it — they put it out and it comes right back in seconds, it was a big flame. So it had to be a big fuel source, it had to be a gas source, because solids don't burn that quickly, won't ignite that quickly. You can go through and analyze it and say what's most probable.
So I asked you about the scientific method. What does it start with? Student: Observation. Yes, collecting the data. And what do you do after you collect it? Student: Create a hypothesis. Very good.
[Tom puts up an NFPA slide.] This is out of the National Fire Protection Association, down here in Quincy, Massachusetts, not-for-profit — the 2014 edition of Guide for Fire and Explosion Investigations. About every three years they publish a new one. It's a consensus document — anyone in the country could write to the committee and say, I think you should rewrite this paragraph. The booklet is about half an inch thick, and every fire investigator uses it. Chapter four is basic methodology — which is the scientific method. Figure 4.3 is the scientific method. Recognize the need — okay, there's a problem. Define the problem, then collect the data around that problem. These four steps are the scientific method. You collect the data, whatever the thing is.
§8. The scientific method at MIT: Nancy Hopkins and the data [47:34]
Let's go back to one I mentioned to you. Nancy Hopkins, a professor of biology, complained to Bob Birgeneau, who was Dean of Science here 20 years ago, that some of the women faculty were not treated equally with the male faculty. Bob Birgeneau was receptive, but he wanted to collect the data. He found out what the salaries of the women were in the School of Science professors, what their office space was in square feet, what was their graduate student space. He had the data, and she was right. So what did he do about it? He corrected it. The next year they all got big raises to bring them up equal to the males in the department.
Chuck Vest was the president then, and MIT got a lot of good press in the Boston Globe. Bob Birgeneau, who went on to be president of the University of Toronto where he came from, and then Chancellor of the University of California system — he's retired now — but he had a lot of integrity. He collected the data, the hypothesis of Nancy Hopkins was that women weren't treated equally, the data showed it was true, and Bob Birgeneau fixed the problem when he drew his conclusion. Then Chuck Vest ordered all the other Deans to do the same thing. You have to ask, why did he have to order them? Couldn't they just by example realize that was a worthwhile thing to do? But no, the president had to order the other Deans, because they already knew what the data would say and they didn't want to know it.
I sat in an engineering council — this is towards the end of my term as department head — and they were under orders of the President to do an evaluation and see if things were unfair in other schools at MIT. They formed a committee, and Professor [Lorna] Gibson was head of the School of Engineering committee that was supposed to look at women's salaries and fairness and lab space. What did she do? She asked Nancy Hopkins, well, how did you go about it? Nancy told her: we collected the hard data. So she made the exact same request to Bob Brown — remember, my great manager that I adore, now president of Boston University. Bob Brown announces to all the department heads that Lorna Gibson's committee has requested salary data, lab space, and all this other stuff, and Bob wanted us to discuss whether we should give it to him.
Well, is that really a question? But we went around the table — ten department heads in the eight departments, because electrical engineering has three department heads. It just so happened I was the last one. We also had an associate dean — the only woman in the room — who didn't generally get to speak unless it was to clarify some administrative thing. Every person in that room before it got to me gave some reason why we should not release the data. This is the truth. It got to me and I said: I don't understand, I don't believe what I just heard. The data is whatever the data is. If you resist this, the president is going to order you to give it to them — so why resist? Why don't you just go collect it? And when you have the data you can figure out what to do with it. In fact that's what they did decide to do. This is not rocket science on my part, this was just simple logic.
But I remember that whole example as how the dynamics of a meeting can have people thinking in ways that make absolutely no sense. That's the true story, folks. When my book comes out it will probably be in the book. Thanks — I see you Monday, I'm lecturing on Monday.