Why Science Needs Failure to Succeed

IRA FLATOW: This is “Science Friday.” I’m Ira Flatow. You know, the narrative about science that receives lots of attention are those big eureka moment, the success stories, the breakthroughs, the cures, the discoveries.

But behind every big innovation there are countless failed projects that are literally swept under the rug. You don’t hear about them. You don’t remember them. Anybody remember the Apple III?

Well my next guest says those failures, the false starts, the dead ends are just as important to the advancement of science. Science could not succeed without the failures.

So what does science lose when you take failure out of the equation? How do we ingrain in teachers and students the value of failure and how to fail better?

Let me introduce my guest. We’re going to be talking about those. Stuart Firestein is a professor of neuroscience at Columbia University. He’s also author of the new book “Failure, Why Science is So Successful.” He’s here with us in our CUNY studio. Welcome back. Always good to see you, Stuart.

STUART FIRESTEIN: Nice to see you Ira. Thanks for having me.

IRA FLATOW: Well let’s talk about that. I want to read one little quote. I was going to ask you to go into the big stuff about it. But then I remember a quote I read. “In science you not only have to have the stomach for failure, you actually have to enjoy the taste of it.”

STUART FIRESTEIN: I couldn’t agree more.

IRA FLATOW: You’re right. Explain that a little more.

STUART FIRESTEIN: Well I think the crucial thing that we all learn, as we become scientists, is that these failures lead us down important paths. We have to be ready to take risks, many of which will fail.

And we can’t be disappointed by these failures. We can’t be disillusioned by them. We can’t be taken off the track by them. What we want is to develop a good taste for failure that takes us to a new place and be open to it.

IRA FLATOW: Give me an example of how failure has helped us succeed.

STUART FIRESTEIN: An example of– well, there are so many of them.

IRA FLATOW: There are so many in the book. Pick one out.

STUART FIRESTEIN: Well let’s say– well, I’ll give you one example is the discovery of a particular kind of a receptor, which was called a G protein receptor. These are very important receptors.

More than 50% of drugs on the market target these receptors. But they were very difficult to find originally, or the part of them was very difficult to find. Because the experiments continued to fail. And it turned out to be it was because of the solution they were washing the glassware in, which contained aluminum fluoride.

And aluminum fluoride is an activator of these receptors, something we had never recognized. And this leads to the whole idea that trace metals are important activators of enzymes. So a failure leads to a brand new discovery that hadn’t even been thought of before.

IRA FLATOW: We’re talking about failures. Our number– 844-724-8255. I think an example that you talk about in the book, that most people know about, is the difference between Isaac Newton and Albert Einstein and gravity. One was a huge failure. Talk about the failure.

STUART FIRESTEIN: Yes, yes. That’s a terrible thing to say. But Newton, of course, had two little things wrong– time and space.

IRA FLATOW: Details, details.

STUART FIRESTEIN: You know, what can you do? But this is one of the remarkable things about science, of course, is that when you say failure, in the common language, you think well, kaput. It’s no good. Just throw it on the heap.

But that’s not true in science. Science can fail for quite a long time and then be updated to something just a bit more correct. And this is really the process by which it works so well, is that it accepts an interim version of what the world is really like.

Now, I mean everything about Newton really works, as long as you stay here on the Earth more or less, or you stay within the solar system and you’re looking at big objects. But if you want to know about you know, the atomic world, and the farther reaches of the universe, the beginning of it– well now you need Einstein.

IRA FLATOW: Why do we never, ever hear about the failures? You know?

STUART FIRESTEIN: Well this is–

IRA FLATOW: We hear, you know, Newton’s again saying if I see further it’s because I’m living on– standing on the shoulders of giants. But we don’t know that maybe those giants have fallen over a couple of times.

STUART FIRESTEIN: Many times. Many times. We have developed, I’m afraid, this kind of heroic narrative about science, this what I call the arc of discovery, where let’s say in physics you go from Copernicus, to Kepler, to Galileo, Newton, Faraday, blah, blah, blah, Einstein, and kaboom. There you have physics.

But that’s not how it went. It took 400 or so years to make that trip. And there were many, many cul-de-sacs and failures and misunderstandings along the way that helped us finally come up with a fairly correct version. But I don’t think there’s a physicist around today who would say we’ve got a complete picture yet.

IRA FLATOW: So what are we losing if we crowd out the failure? What are we leaving out that is useful to know?

STUART FIRESTEIN: Well we’re leaving out, I think, about 90% of science, is the problem. And we have begun to crowd out the failure. We don’t teach it in schools. We teach textbook versions of science, these arcs of discovery.

We don’t talk about the cul-de-sacs, or the mistakes, the wrong turns, from which we learned a great deal. Embryology is a very good example of that. The whole field of embryology, for a long time, was mistaken about things. But it sort of– phrenology is even better, since I’m a neuroscientist– you know, the bumps on your head being something about your personality.

That was a completely bankrupt idea. But for 50 years it operated as a science. And it still is– has foundational ideas that are important in modern day neuroscience.

IRA FLATOW: Could we teach intelligent design as a failing?

STUART FIRESTEIN: Yes, I think– I happen to believe that we should teach intelligent design in classrooms. I think it’s a perfectly reasonable thing to teach. Let’s remember that Newton, Faraday, many great scientists, clearly would have believed in intelligent design.

So I think we should teach it and understand it as, in my opinion of course, a failure along the way, but not entirely. There are even things in it that are probably acceptable to evolutionary theory and could be wedded into it. And we should understand how we get from that to evolution.

IRA FLATOW: I want to bring on another guest to talk more about this, a school teacher who teaches this idea of failure to her middle school students. Helen Snodgrass is the science director of YES Prep Schools and an AP Biology teacher– I’d ever get into that class– at YES Prep North Forest School in Houston.

Welcome to “Science Friday.”

HELEN SNODGRASS: Hi, Ira. Thanks so much for having me on.

IRA FLATOW: You’re welcome. On your chalkboard, I understand– on your chalkboard– boy, chalkboard, you have written out failure is not an option. It’s a requirement. Why do you say that?

HELEN SNODGRASS: Yes, I do have that up there. It is definitely a quote that takes students aback a little bit when they first come into my classroom. But I have that quote up there for all of the reasons you all have been talking about.

Students often come into science class with this idea that failure is wrong. Not knowing the answer right away is wrong. Doing an experiment and getting an unexpected result is wrong.

And we know that that is so different from how the actual process of science works. So I use that to sort of catch their eye right away and sort of launch us into that year-long conversation of the real process of science.

IRA FLATOW: Is it– is it– does it relieve them to know that they can fail and that takes some of the burden off and make them better students?

HELEN SNODGRASS: Yeah, I think it’s initially a little scary. Because it’s not what you’re taught in school, that failure’s OK. But I think over time it actually does relieve them.

I actually had a– I talked to my students this morning, since I was thinking about this discussion, to sort of find out if their views had changed over the course of this year. And I had one student who said you know, I really used to be embarrassed when I got something wrong or I didn’t know the answer.

And now I actually feel relieved and I understand that I’m not expected to know all the answers in any given moment. And I get to ask questions. And I get to try something a different way.

IRA FLATOW: Stuart, I see you’re nodding your head in agreement, and at the college level also.

STUART FIRESTEIN: Certainly at the college level. And Helen, let me congratulate you. It’s a brilliant essay that you posted in “The Washington Post.”

HELEN SNODGRASS: Oh, thank you.

STUART FIRESTEIN: I just think it’s exactly right in so many ways. I wish I were taking your class, or had taken your class. And I think this is a critical thing.

Because this notion of failure being acceptable, and indeed being a requirement for the process, an integral part of the process, most of us don’t get until we reach frankly graduate school, when we’re going to become professional scientists.

That’s when you learn that the answers don’t really count for much. It’s the questions you ask. And that you fail a lot, but that you have to learn things from the failures. And it’s unfortunate.

Because for most of the population, I’m afraid, the last interaction they have with science is either a high school class or an early college class, that’s just based on the accumulation of facts and memorizing a bunch of stuff. So Helen should run the curriculum, if you ask me.

IRA FLATOW: Well I see–

HELEN SNODGRASS: I don’t know about that, but–

IRA FLATOW: Let’s see what our listeners have to say.

STUART FIRESTEIN: I’m not sure she wants the job, but–

IRA FLATOW: Talking with Stuart Firestein, author of “Failure, Why Science is So Successful,” and also with Helen Snodgrass, who teaches failure in her school. Our number– 844-724-8255, if you’d like to join us. You can also tweet us @scifri.

How do you bring it in, Helen? How do you teach it in the classroom? How do you instill that’s OK?

HELEN SNODGRASS: Yeah, so I mean, I think the first thing is just introducing students to the idea that struggle and failure are both normal and essential parts of science. Because they’re not going to come in with that idea.

You know, one way I do that– with that sort of provocative quote. And I also you know, give them some examples of things that were initially perceived as failures but ended up actually leading to something new and interesting. And I’m really excited about Dr. Firestein’s book, because I think I’ll get a whole bunch of other examples for my class.

IRA FLATOW: Yeah.

HELEN SNODGRASS: But another thing is just giving students the opportunity to struggle and fail by you know, giving them more open-ended tasks and more open-ended labs to do, where there isn’t necessarily one clear answer, or one right answer.

IRA FLATOW: But you know, our whole culture is built around success. I mean, no one remembers who lost the Super Bowl or the– or the World Series. Well, I’ll remember the World Series.

STUART FIRESTEIN: Yeah.

IRA FLATOW: But you know, we don’t think– we don’t think that second best, or failure in those senses, is– teaches us anything from it.

STUART FIRESTEIN: I don’t think we’re talking about failure necessarily in those senses. I mean, there are failures that aren’t good. I can’t suggest that everybody should just fail away and that’ll be just fine, of course.

And it’s not even a question of failing in order to become successful later on. It’s the fact that if you embrace failure as part of the process– not the whole process, but a part of the process– then this is the place where you can be most creative, most innovative, where the most interesting things happen, I think.

IRA FLATOW: Let’s see if we can get a couple of calls in from our listeners. Let’s go to Stu in Grand Blanc, Michigan. Hi, Stu.

CALLER: Hello. I’m very interested in this topic. I’m a retired anthropology professor. And over the years I’ve taught a little bit of history of science in my intro classes.

And one of the most important experiments in the history of physics, for example, was the Michelson-Morley experiment. And I frequently ask my physics colleagues why don’t you teach that? And they say oh, it was a failure. And I said yes, but it set the groundwork for relativity and for quantum mechanics.

IRA FLATOW: Great point.

CALLER: So I was–

IRA FLATOW: That’s a great–

CALLER: I was wondering what– what Stuart Firestein had to say Michelson-Morley.

IRA FLATOW: OK, thanks.

STUART FIRESTEIN: That– that’s often quoted as, of course, the classic version of how a failed experiment– a so-called failed experiment– an attempt to measure the ether and light as it sped through the ether, gave rise to a kind– almost kind of a crisis in physics and finally to relativity.

But it– but that sort of thing happens, I have to say, again and again in science, that– lets– another example would be the one that I think we should still teach is caloric, the idea that heat flows, that it’s a substance that flows.

Now we know now that that’s not true. Heat is actually motion. It’s the jiggling of atoms. But to get from the idea of caloric– because really, heat does seem like it flows. You put a hot thing next to a cold thing–

IRA FLATOW: Right.

STUART FIRESTEIN: –and you could measure it as a flow. But it’s not really a flow. And the idea that it’s motion is the basis of thermodynamics, of subject that sounds so frightening to teach we would never try it.

But it’s really relatively easy to teach, I think, if you start from where we were wrong and see how we got to be right.

IRA FLATOW: Helen, with such emphasis on testing and state exams, how do you balance prepping students for what they need to know and then the idea of failure?

HELEN SNODGRASS: Yeah, I think that definitely presents a challenge to teachers. And I would say a few things to them, people concerned about that, which is one is that easy learning, the sort of learning we think of as you know, test prep for a high stakes test– it doesn’t actually result in very deep or long-lasting learning.

So learning then involves more struggle. And you know, more failure might initially not feel as good, but is actually going to lead to students understanding more in the long term, which we hope will show up on those assessments.

The other thing I would say is that every teacher knows that students who aren’t very engaged in your class are not going to do as well. So if you’re presenting science as this long list of facts, a lot of students aren’t going to be as interested, or as invested in that.

IRA FLATOW: Yeah. Well, I want to wish you good luck, Helen. Thank– thank you for taking time to be with us today.

HELEN SNODGRASS: Yeah, thank you so much for having me.

IRA FLATOW: Helen Snodgrass- science director for YES prep schools and an AP Biology teacher at YES Prep North Forest School in Houston, Texas. I’m Ira Flatow. This is “Science Friday” from PRI, Public Radio International.

Still talking about failure with the author of the book “Failure, Why Science is So Successful,” Stuart Firestein. Do you think we can turn around and get people to understand you know, that failure– I mean, certainly in this political science, right–

STUART FIRESTEIN: Yes.

IRA FLATOW: No one wants to be looked at as being a failure or failing at anything.

STUART FIRESTEIN: Yeah, I think one of the worst things– I don’t want to talk about politics.

IRA FLATOW: Yeah, good.

STUART FIRESTEIN: For sure. But– but I think one of the worst charges we make about politicians is that if they change their mind, their flip floppers. That’s a terrible thing.

We all should change our mind. If the facts change, you should change your opinion, you know, has been said. And so I have no problem with people who revise their opinions about things, or revise their ideas about things.

One of my favorite quotes about science is that in science revision is a victory. We’re always revising.

IRA FLATOW: Mm hm. Let’s even get another phone call before the break. Eric in Gainesville, Florida. Hi, Eric.

CALLER: Hey, how are you doing?

IRA FLATOW: Hey there. Go ahead.

CALLER: So I’m a graduate student in a physics program. And the competition is really hard– or, really tough. And I learn by failure. So like, I do– if I– that’s how I’ll remember something for a long time. And I feel like it’s really tough, in a graduate program, where you’re expected not to fail.

STUART FIRESTEIN: I’m not sure what graduate program you’re in. But the graduate programs I know I think put up– one of the things we do is we put up with a great deal of failure.

We– at least I, certainly. I run a laboratory, here at Columbia. And we study the sense of olfaction as part of a neuroscience project. And I anticipate that my students will fail quite a bit.

In fact, I have to say my experience, and one that I’ve shared with many colleagues, is that almost all the data that winds up in a thesis happens in the last 18 or so months of a graduate career that typically lasts five or six years– pardon me.

Now it may be different in physics. I’m not quite sure of that. But in biology I think we have a pretty high sense of failure that’s OK for us. We have a very good tolerance for it.

IRA FLATOW: Yeah, and in– the world of invention is full of failure. Right?

STUART FIRESTEIN: Edison famously said I found 10,000 ways to fail before the light bulb popped on. You know.

IRA FLATOW: My favorite invention of a failure is Silly Putty. Silly Putty was supposed to be a substitute for rubber.

STUART FIRESTEIN: Yep, yep.

IRA FLATOW: And it was– one with it– it would dissolve in water. But it became a great toy.

STUART FIRESTEIN: Yes. Post-it notes, you know– Post-it notes were a failed adhesive. I mean, they don’t really stick, right? That’s the problem.

IRA FLATOW: Made a lot of money with that–

STUART FIRESTEIN: Yes, pretty good, huh?

IRA FLATOW: –failure– failures. And that’s the thing, is that you really don’t know where a lot of these things are going to be leading to. They open the door. One thing is closed opens the door to another idea.

STUART FIRESTEIN: I think that’s the crucial notion here, that we want– if we want to be creative in science, if we want science to remain quote successful, then we have to leave those doors open. We have to know where those places are.

IRA FLATOW: When we come– when the government puts out grants– NIH, National Science Foundation– do they allow for failure in the experiment?

STUART FIRESTEIN: Well, this is one of my rants, if you will, of course.

IRA FLATOW: You got a minute to rant before we–

STUART FIRESTEIN: OK, a quick minute to rant. I mean, I think they don’t. I think the funding levels are too low to put up with failure. And so grants become sales pitches.

Famously NIH and the NSF, the other major funding group, have what are called high risk, high impact grants. So these are risky grants. But they’ll have a great impact.

Now last year that was 1.5% of the NIH budget. So that’s fine. But what does that mean? That 98.5% of the budget is going to incremental stuff that we don’t really care about and don’t expect that much of an impact? That’s not really the way we want to spend our money, is it?

IRA FLATOW: No. Well, we’ll talk more about it with a Stuart Firestein, author of “Failure, Why Science is So Successful.” 844-724-8255 is our number. You can also tweet us @scifri. We’ll take a short break. See you on the other side of it.

This is “Science Friday.” I’m Ira Flatow. We’re talking this hour about the importance of failure in science. My guest is neuroscientist Stuart Firestein, author of the new book “Failure, Why Science is So Successful,” on heels of your previous book, which was–

STUART FIRESTEIN: “Ignorance, How Science– What Drives Science.” Yes, ignorance and failure. I’m really carving out a little niche for myself here, I think.

IRA FLATOW: But there is a sort of a link between the two, right?

STUART FIRESTEIN: Well, I think there’s a crucial link between the two. I mean, first of all, ignorance, again, sometimes used in a pejorative sense, but I mean it in the better sense, that is the what we don’t know in the world.

And that’s what science is about finding out. But– but then there’s, as Donald Rumsfeld famously told us, the what we don’t know we don’t know. And that’s what you find out by failing, really. So I think failure ties into ignorance by actually deepening it.

IRA FLATOW: All right, let’s see if we can go to the phones. We’ve got some pretty good phone calls here. Let’s pick them up. Let’s go to Peter in Chicago. Hi, Peter. Woop.

CALLER: Hello?

IRA FLATOW: Yeah, go ahead.

CALLER: It’s Raleigh– Mark Anderson.

IRA FLATOW: I’m sorry, go ahead.

CALLER: Hi, am I on the air?

IRA FLATOW: Yes.

CALLER: Hi. I’ve been listening with great interest. I’m a pharmaceutical scientist. I’ve been practicing now in the consulting area for 10 years. And I recall from my graduate school days a very wise postdoc mentor told me that you have to measure success in terms of six month increments.

And I think the reason is, of course, is because most of your time is spent having these so-called failures. What I would also like to add, though, is as a practicing scientist over the years I’ve had time to sort of organize things in my mind a bit– maybe not quite as refined as the author here.

But what I find is that a failed experiment is one where the data is not good. So I think so long as you have reproducible data, if it doesn’t answer the question that you’ve asked, it refines your thinking, makes you ask a different question, and ultimately leads to the truth.

So I think I just wanted to add that point, that you know, to me a failure in my job is when I can’t do anything with the data. And anyway, I’ve been listening to this with great interest. Thank you for the time.

IRA FLATOW: All right. Thank you. You had some interesting stuff about data, talking about failure in data.

STUART FIRESTEIN: Yes. So I mean, I think the caller is correct in that failures that are just due to sloppiness, or poor data, or not taking care of things of that nature– they’re not interesting.

No, there are uninteresting failures. And there are interesting failures. Of course. There are uninteresting successes, frankly.

I mean some facts are– oh, that’s great. Put that away. And let’s get on with it. So it’s the interesting failures that are the ones that are important. And I’m not talking about those that are just because of poor technique or something like that.

I mean, there’s a wonderful quote from Isaac Asimov who once said that what every scientist wants to hear most when they see new data is not the word eureka, but oh, that’s weird. Because that tells you there’s something yet more that you haven’t really surrounded yet, somehow, you know?

IRA FLATOW: Now let’s talk a little bit about how we teach science to kids in school.

STUART FIRESTEIN: Yeah.

IRA FLATOW: This is– it’s my soapbox every once in a while. And I think that we don’t teach science enough about how science is carried out. We teach you how to titrate some liquid til the color changes, you know?

STUART FIRESTEIN: Yep.

IRA FLATOW: But we don’t teach you why science is important, or, in the larger sense that we’re talking about, why failure is important in science.

STUART FIRESTEIN: Yes, I think we leave out the questions. And we leave out the failures, and thereby leave out almost all the important stuff in science. And worse than that, we actually, I think, provide a distorted view of science, that much of the public then has, which that it’s a big book of facts.

That it’s set in stone. That it’s irrevocable. Worst of all, that it’s the truth with a capital T, none of which, I think, any scientist would make any claim towards, at all. But that is what we teach. And you’re right.

Now I personally think the problem there is one of evaluation and assessment. I think it can be fixed. And I think it could be fixed quickly, if we did some hard work on developing new tools, more refined tools, of evaluation and assessment, things better than multiple choice exams, for example, that really only memor– measure memorization.

IRA FLATOW: There’s a lot more good stuff coming from Stuart in his book “Failure, Why Science is So Successful.” Stuart Firestein– welcome back. Good to see you.

STUART FIRESTEIN: Thank you so much, Ira.

IRA FLATOW: Hope you’re working on the next one.

STUART FIRESTEIN: I– thinking about the next one. It’s always a pleasure to be here. I have a wonderful audience with great questions.

IRA FLATOW: Thank you. And you actually can read an excerpt of his book on our website. It’s sciencefriday.com/failure.

Copyright © 2015 Science Friday Initiative. All rights reserved. Science Friday transcripts are produced on a tight deadline by 3Play Media. Fidelity to the original aired/published audio or video file might vary, and text might be updated or amended in the future. For the authoritative record of ScienceFriday’s programming, please visit the original aired/published recording. For terms of use and more information, visit our polices pages at http://www.sciencefriday.com/about/policies.