Join us April 17-19 for the BioLogos national conference, Faith & Science 2024, as we explore God’s Word and God’s World together!

Forums
Featuring guest Larry Molnar

A Wrinkle Occurred | The Nature of Science with Larry Molnar

A story of an astronomer and a big scientific claim highlights the ways that science leads to knowledge about the world, and not always in straightforward ways.


Share  
Twitter
Facebook
LinkedIn
Print
16 Comments
16 Comments
Silhouette of person looking at stars

A story of an astronomer and a big scientific claim highlights the ways that science leads to knowledge about the world, and not always in straightforward ways.

Description

Larry Molnar spent years with his astronomy students studying a contact binary star system. Their observations and data eventually led them to make the claim that the stars were headed toward an explosion called a luminous red nova, a major discovery. The story of these stars and of the scientists who study them highlight the ways that science leads to knowledge about the world, and not always in straightforward ways.

Subscribe to the podcast


Transcript

Hoogerwerf: 

Welcome to Language of God. I’m Colin Hoogerwerf. I’m here in the studio today with Faith Stults. 

Stults:

Hey Colin. 

Hoogerwerf: 

Hey. So, I’m going to let you lead this one today, because you initially brought this story to my attention. 

Stults:

Yeah, it’s a story that shows, in some really interesting ways, how science works, how it brings about knowledge, and how it’s not always straightforward. 

Hoogerwerf: 

Should we just jump right in? 

Stults:

Yeah, let’s go. And our story takes place in space…

Hoogerwerf: 

With a couple of stars…

Stults:

And an astronomer…    

Hoogerwerf: 

And eventually leads to published papers and press conferences around a pretty major scientific claim.

Molnar: 

Yeah, my name is professor Larry Molnar and I teach at Calvin University. 

Stults:

Larry started at Calvin back in 1998 when they started the astronomy program. And he was actually hired alongside Deb Haarsma, who now, of course, is the President here at BioLogos. 

Molnar: 

The goal was to start astronomy and not to have a lonely astronomer, so they should hire two.

Hoogerwerf: 

Larry had come to Calvin from doing graduate level research of a kind that was probably not appropriate for undergrads, so his main question when he started was how to do research in a way that would engage his students.  

Molnar: 

The main method I used for taking data in astronomy before that was radio astronomy, which is a wonderful field. And you can make images through something called interferometry, which is a little bit daunting to a first year undergraduate student.

Stults:

So, instead, they went about building a small observatory, with one telescope in New Mexico and another in Grand Rapids, where Calvin is located. 

Molnar: 

And our goal was to use that not only in classes, but in research

Stults:

And pretty soon they were off and running. It wasn’t a huge telescope, but there are actually some advantages to that. Most research telescopes are shared by astronomers from many institutions and even countries. So if you want to use it, you have to submit a request months in advance, get it approved, and then you’re only given a few short nights of telescope time. And if it’s cloudy on your assigned nights… tough luck. But Calvin didn’t share it with anyone. So if a student had a question or a project that interested them, they could immediately just go to the telescope, point it wherever they liked, and take as many images as they wanted. 

Hoogerwerf: 

And that’s exactly what some of his students started to do. 

Molnar: 

The most impressive undergraduate research that I had participated with was actually a student, Melissa Dykhuis, who spent hundreds of nights looking at asteroids

Stults:

And when you’re studying asteroids, you end up taking *a lot* of pictures of the whole sky. 

Molnar: 

Then I had another student, Dan Van Noord, who was a high school student coming to me saying, I’m going to go to Calvin. And I’m going to study binary stars when I get there. And just letting you know. So when he came, he asked what we had done with the telescope, and I had described to him Melissa’s work. And he said, “perfect.” She has hundreds of nights where she looked all night long in one direction to study a particular asteroid. But each of those nights in a different direction. He systematically looked through all of those nights of data, not at the asteroid—he didn’t care about asteroids—but at all the stars in the field to find, “Are there any variable stars, not yet known?” And he wrote programs to process all of that and discovered dozens of binary stars, including contact binary stars in the process.

Hoogerwerf: 

Ok, let’s pause for a minute, because we’re getting into some real astronomy here. And it turns out your background, Faith, is in astronomy. So here’s the first of what might be many astronomy questions. A contact binary? I can maybe guess that a binary star is two stars?

Stults:

Exactly. Even though our Sun is a lone star hanging out in the Milky Way by itself, astronomers actually think that most stars form in pairs or maybe groups. This happens when a large cloud of dust and gas starts contracting because of its own gravity. And more often than not, that cloud – or nebula – will fragment into a couple smaller clumps, each of which then go on to become stars. Since these stars formed so close to each other, their mutual gravity continues to draw them together and they wind up orbiting each other for their entire lives.

Hoogerwerf: 

Wait, does that mean that the twin suns of Tatooine are scientifically accurate??

Stults:

Oh you bet.

Hoogerwerf: 

Alright, good to know. And contact, meaning the stars are touching?

Stults:

Precisely. Unlike planets that have discrete solid surfaces, stars are basically balls of gas that are incredibly dense at the cores and then get more and more diffuse as you move outward. So when we look at the Sun in the sky, it looks like it has an edge, but actually there’s an atmosphere that extends beyond that. So a contact binary system is when two stars are orbiting so close that their atmospheres overlap with each other. They may even be passing material back and forth. 

Hoogerwerf: 

So do we call this a star or stars or a system?

Stults:

Yeah good question. Technically there are multiple stars there, so calling it a star system is probably most accurate. But when you’re looking at something like this through a telescope, it just looks like one point of light, no matter how big your telescope is. So often when astronomers first discover these systems, they assume it’s only a single star, only to realize later that there are actually 2 or 3 stars present. So binary star system, binary star, binary system… it all means the same thing

Hoogerwerf: 

Ok, continuing with the story. Larry’s student, Dan, has discovered a bunch of these new binary systems. What happens next?

Stults:

Yeah, so Dan keeps looking at these stars and reporting new ones to the official catalog. 

Molnar: 

Dan learned how to take data, how to process that data, how to recognize what kind of star you have, and in particular, how to get very precise orbital periods.

Stults:

So an orbital period is just how long something takes to complete an orbit. And a binary system’s orbital period is one of its most important characteristics. Learning how to do those calculations turned out to be really well timed because a little later on, a conference was held, not too far from Calvin, that Larry and some of his students planned to go to. And before they went they read the abstracts for some of the presentations.

Molnar: 

And one abstract was very interesting in that it said, there’s this very strange star, KIC9832227. 

Hoogerwerf: 

And the presenter didn’t know what kind of star this was? 

Stults:

Yeah. Her particular focus was in pulsing or variable stars—stars that expand and contract—getting brighter and dimmer in the process. And so once again, we have to remember that when an astronomer is looking through a telescope, they’re still only seeing a pinpoint of light. So you’re making all kinds of inferences about what a star is like by changes in its light. So this star, KIC9832227, didn’t have a constant brightness. 

Hoogerwerf: 

Which is what you’d expect for a variable star?

Stults:

Exactly. But it was also possible that the brightness was changing because it was actually a binary star: two stars orbiting each other, with one sometimes blocking the other’s light. And she wasn’t sure which of those things was happening. 

Hoogerwerf: 

But Dan knew. 

Stults:

Yup, after reading the abstract he took that as a challenge and he went and took some of his own data with Calvin’s telescope. 

Hoogerwerf: 

And so off they went to the conference, Dan, with his data in hand, where this presenter gave her presentation on KIC9832227. 

Molnar: 

So after she gave her presentation and said, “We need to figure out what kind of star this is—” 

Stults:

And I love this part, right after she gives her presentation with everyone there in the audience at this astronomy conference…

Molnar: 

—Dan raised his hand and said, “I know what kind of star this is,” and showed her the evidence that indeed it was a binary star.

Stults:

So that was pretty cool. And it kind of jump started Dan’s interest in this star. 

Molnar: 

And he was very industrious as far as going back to the literature to see who else might have observed this in some catalog. And he’s extremely good with computers, so downloading any data he could find from various sources. And so he took the data that this woman had, which was from the Kepler satellite and determined the best orbital period from that. He took data from previous ground base surveys and determined the period from that. And then he found something that perplexed him. He literally came to me and said, “What have I done wrong? I didn’t get the right answer.”

Stults:

This is an important point in the story. Up until now, the science has been pretty smooth. Everything has fit into place pretty well. The scientific method was running like a well oiled machine. Beginning with a question:

Hoogerwerf: 

What kind of star is it? 

Stults:

A hypothesis was put forward.

Hoogerwerf: 

It’s a contact binary system. 

Stults:

Observations were made, literature was reviewed. Data was analyzed.

Hoogerwerf: 

And those observations and calculations seemed to support the hypothesis. KIC9832227 is a contact binary. 

Stults:

And supporting your hypothesis is always nice. But occasionally we do the testing, and the results surprise us. In this case, the data supported the contact binary hypothesis, but it also suggested that the rate at which the two stars were orbiting each other was getting shorter, which was not what they expected. 

Hoogerwerf: 

Why not? 

Stults:

Well, typically the orbits of binary stars stay pretty stable for the stars’ lifetimes, say a couple million years. So you wouldn’t necessarily expect to see the orbits shrinking quickly over the course of a couple years like Dan’s star appeared to be.

Hoogerwerf: 

And when unexpected data comes in it can mean a lot of things. It could mean that you just did something wrong. It could mean that you overlooked a possibility. Or it could mean that something else is happening that we didn’t understand completely before and we need to refine our hypothesis. 

Molnar: 

This is what I love as far as students learning about science. It’s not about getting the answer that’s in the book. It’s not even about getting the answer that the professor has in his head. It’s about getting the answer that nature has. And at this point, I knew Dan understood how to get these error bars, that he knew how to get the periods. So I said, “You’ve done nothing wrong, you know, you’ve done nothing wrong. You’re doing exactly what you’ve done before. You’ve just proven that the period changed.” And then, he immediately changed, was like, okay, let’s not let’s find out what I did wrong. Let’s find out what’s happening.

Hoogerwerf: 

So what’s happening?

Stults:

So at this point, Dan had already figured out that this star was a contact binary system, two stars orbiting each other really closely. But what he started to see is that not only were they orbiting each other, but the orbit seemed to be getting smaller. 

Hoogerwerf: 

And so another scientist comes in here, Romuald Tylenda, from Poland…

Stults:

Who had also been looking at contact binaries and had done some research on a binary system that, in 2008, came to the end of its life, which happens to be pretty exciting.  

Molnar: 

The end of the stars is an amazing explosion, so called a luminous red nova explosion.

Hoogerwerf: 

So this is not a supernova, not that I’m sure I even really know what that is. Let’s take another astronomy break. 

Stults:

Yeah, not a supernova. So in astronomy there are a number of different types of novas. The word nova means ‘new’, and this term comes from many centuries ago when astronomers would suddenly notice a new very bright light in the sky, what looked like a new star. 

Hoogerwerf: 

But actually it’s not new.

Stults:

Almost the opposite. It’s the death of a star. Of course they didn’t know that, it’s just that the death of a star gets very bright, so a star they might not have seen or noticed with the naked eye, would all of a sudden become shockingly bright. And there’s lots of different kinds of stars, different sizes and densities, and they have different ways they end their life. And so a supernova is how one type of star dies that happens to be incredibly bright and dramatic. A luminous red nova is another. A little less dramatic visually, but still very cool.

Hoogerwerf: 

Ok, good enough for me. So back to our stars.

Stults:

Yeah, Tylenda had seen his binary system only after the Red Nova had already happened. 

Hoogerwerf: 

And these red novas aren’t some incredibly rare event. 

Stults:

No, they’re not incredibly rare, at least by astronomy standards. Maybe around one a decade in our galaxy. So it’s not necessarily common either. But as it happens there are a lot of stars in the sky. And the sky is really pretty big. So an event that happens once in a decade, with no warning, and that needs to be caught by a telescope… let’s just say you’d have to be really lucky to watch it happen. So the only way anyone has ever witnessed a red nova is by seeing just the aftermath of one.

Hoogerwerf: 

Like Tylenda did. 

Molnar: 

But that’s not the same as a targeted observation. It’s never been done before, because nobody ever had a plan, a way to identify a thing that’s about to explode.

Stults:

Tylenda’s work helped to show what the orbits of these stars might be doing in the lead up to a red nova. So then the obvious question that came up was, could the changes that Tylenda saw in the star in 2008 be analogous to the changes that were happening in KIC9832227? 

Molnar: 

We both agreed, not likely. But we have a telescope, we can just start watching it. We literally scheduled observations for it for the whole of the fall and the spring semester, and didn’t really look at them. Because we had classes and whatnot. We saved them up for the following summer. Because it wasn’t likely anything dramatic was going to happen. But when we looked at it that summer, we realized between the time of the Kepler spacecraft data and our own data, it had speeded up yet more. And so hey, this is a real possibility now.

Stults:

So here we have science at work again. The data and observations that were made led to a new hypothesis.

Hoogerwerf: 

That this contact binary star was heading toward an explosion. 

Stults:

And so they go and they take a bunch of new pictures, gather a bunch of new data, and it continues to support that hypothesis. But, as I think is becoming clear, most science doesn’t work the way that we learn about as kids, where we have a hypothesis, we do one experiment and what it shows us proves once and for all the way the world works. 

Hoogerwerf: 

In actuality, science is more cyclical. 

Stults:

Or I kind of like the word iterative.

Hoogerwerf: 

What we learn presents new questions, new data asks us to refine our models of the way things work. And in the case of Larry and Dan’s star, their observations had led them to understand that it was a contact binary star, that its orbital period was decreasing, and that it was doing so in a way that looked analogous to the 2008 star which ended in a red nova. So the new hypothesis was that this star would also end in a red nova. 

Stults:

And soon. 

Hoogerwerf: 

Right. But we don’t end there because science doesn’t prove things in that way.  

Molnar: 

The language I use in my classes is the language of models. And you can say the true model is what really is happening. False model’s something you’ve proven isn’t happening. But having said that, you can never prove something’s true. Therefore, that’s not really the useful adjective. The useful adjective, in the one I talk about is a powerful model. A powerful model is consistent with what we know so far, and makes further predictions that we can check. So that’s its power. And if it matches those predictions, that maybe gives you more confidence in it. But it never goes into the category of true, it’s always just more and more powerful. 

Faith:

I think this is a really good way to think about science. 

Hoogerwerf: 

Yeah, and I’ve always thought the example of gravity is really good here. I think non-scientists think about gravity as something that is just purely proven. We did the work, we came up with the answer. As Bill Nye used to say, “And now you know.” 

Stults:

But gravity is still just a model. It’s a really really powerful model. At least here on Earth. But there’s actually a lot of problems with it when you start thinking about black holes and more advanced physics. 

Hoogerwerf: 

Another example of a model like this is the science of climate change. And maybe this has kind of the opposite perception, which is that a lot of people think that we still don’t know a lot, that climate change is a weak scientific model, when in fact most scientists would consider it a very powerful model and the predictions it makes continue to be accurate. There’s a really interesting visual way of thinking about this. 

Stults:

You are aware this is a podcast, right? No, no visuals.  

Hoogerwerf: 

Yes, Thanks Faith. I am aware, but I think our listeners have a good enough imagination to follow along with me here: Picture a white board. Or a chalkboard if you’d rather. And then imagine placing four dots in a grid, in what could be the corners of a square. The dots are data. The shape you can make with the dots is the model. When you look at the four dots, you think square. The data tells me this is a square. 

Stults:

Alright, I’m with you so far. 

Hoogerwerf: 

So you draw the line connecting the dots into a square. Well instead of stopping there, the scientist says, why don’t we get more data. Place another dot, your new data, this one in between the top two dots, but just a bit above the line to your square. All of a sudden your model isn’t very powerful because it can’t explain all the dots. But you step back and realize that a circle will connect all the dots. 

Stults:

So a circle is your new model? 

Hoogerwerf: 

Right. And you gather more data, place a bunch more dots and they all fit perfectly onto different spots on the circle. A circle becomes a more powerful model. 

Stults:

So bringing this back to Dan and Larry. They’ve built what starts to feel like a pretty powerful model for this binary star that looks like it might be heading toward an explosion in the very near future. All their data is consistently fitting the shape of their model. And if they’re right… well let’s just say it would be a really big deal. 

Molnar: 

What we need to really work on now is what are the other possibilities? And can we rule out all other possibilities? And that’s what took up our time following that.

Hoogerwerf: 

So what were some of the other possibilities? 

Stults:

So there could be several possibilities. Maybe even an infinite amount of possibilities. But you can only evaluate the ones you can actually think of. So one of those is that there could be a third star out there affecting things or tugging on the orbits.

Hoogerwerf: 

And that’s what Larry called a third body.

Stults:

Yup. And that’s one possibility that he took a lot of time to rule out. But eventually you get to a point when your model looks good enough that you can’t keep it a secret. You’ve got to tell the world about it. 

Hoogerwerf: 

So they’ve done all the work, written the article, and they’re ready to go out and make this big claim that a star is going to blow up in 5 years. They’ve got a press conference scheduled and everything. 

Molnar: 

Yes, Eventually you have to do that.

Stults:

And then there’s a detour. 

Hoogerwerf: 

What happened?

Stults:

He realized he wasn’t totally sure about the third body thing. And maybe he realized with the mounting pressure that he needed to really button down and get a little bit more supporting data first.  

Molnar: 

So literally, the day before that press conference, I called the guy who’s organizing the press conference and I said, “we’re going to have to come back to this.” And it took us another couple years before we got the spectra that absolutely verify that it was not the third body and said, “We’re ready now for the press conference.” And so 2017, we finally did the press conference, and wrote the paper and got some publicity about it.

Hoogerwerf: 

So they did it? 

Stults:

Yeah!

Hoogerwerf: 

And was it super exciting? 

Stults:

Yeah! Larry’s being pretty humble here, but his prediction definitely made a splash in the astronomical community. I was actually teaching high school physics and astronomy at the time, and the following day I actually started class by excitedly telling them all about this. This was the first time an astronomer precisely predicted a nova, which would allow other astronomers to actually watch the explosion as it unfolded for the first time ever. 

Hoogerwerf: 

And how did the scientific community respond?

Molnar: 

So, two different kinds of responses. On the one hand, many people were excited about it, and thought, what ways can we observe to begin to take advantage of this or maybe add to the story in some way. The other side was some European astronomers thought it just unlikely, and wrote as much in certain papers saying, we thought about this, but we dismiss it. What’s striking to me though, when it comes to powerful models, when it comes to the testing and science, is if you want to be dismissive, you need to have a reason. We laid out the case against all of the alternatives we could come up with. Those who were dismissive in that first year, had no alternative. And so I was dismissive of the dismissive folk. Because it’s simply not good enough to say “I don’t think it’s likely.” If you think it’s wrong, you have to say what do you think instead? And that was not happening. So for the next year, while there was push back, it was not push back that I considered powerful science. It wasn’t push back that made any prediction, it wasn’t push back that I could address, because it was just “I don’t think so.” At the same time, the star was continuing to follow our prediction so I couldn’t change it anymore. It would either follow or didn’t follow what we said. And it followed it. So a year on, I was still very excited, in fact, more excited than ever about it.

Stults:

So there were the dismissive people. But there were also those first kind, the kind that were excited. But their excitement and support took the form of, “Hey, this is really interesting, let’s check it out. And let’s get some more data.” 

Molnar: 

And it was precisely from those supportive people that a wrinkle occurred. 

Hoogerwerf: 

Uh oh.

Stults:

Yeah. So another group of scientists, led by a graduate student at San Diego State University had a data set that hadn’t been published yet and that filled in an earlier gap in Larry’s model.

Hoogerwerf: 

So if the orbital period is speeding up over time, if you look backwards you should be able to see that it was slower and Larry’s model should tell you just how slow. 

Stults:

Right, and so this group analyzed this new data and compared it to Larry’s model. And they found that it didn’t quite match. 

Molnar: 

Within the uncertainties, it was significantly off. So that made them think “Oh, maybe Molnar doesn’t know what he’s doing.” And they re-analyzed all of the epochs that we had analyzed, and found that everything agreed, except the very first data point. So in addition to the one they had, there was now a disagreement about a 1999 data point where we found a certain timing, and they found a different timing.

Hoogerwerf: 

So first they have this point from a data set Larry didn’t have. And it doesn’t fall onto the place it should in Larry’s model. And then they go back and look at all of the data points Larry did have and they find a problem with one of those too? 

Stults:

Yeah. So this 1999 data point is from a previously published data set called the Northern Sky Variability Survey. Stick with me here. And it’s one that Larry had used in his initial paper, but when this group did the analysis they got different results than Larry did. And they started putting together a paper that rejected Larry’s hypothesis that the stars were heading toward a red nova in 5 years. And then they sent it in for peer review. 

Hoogerwerf:

Let’s take a minute to talk about peer review, because it’s really important for the way science works and it plays an important role in this story.

Stults:

Yeah so science is fundamentally a community endeavor and only works when a lot of people are sharing ideas and challenging each other’s ideas. 

Hoogerwerf: 

And journals and articles are the main way that happens. 

Stults:

And so the idea is that when you want to share a finding from your research with the rest of the scientific community, you publish it in a journal for everyone to read. But before it is actually published, it goes through a process where it gets sent out to a bunch of other scientists who get to give feedback. 

Hoogerwerf: 

And they call those other scientist reviewers referees.

Stults:

And so by the time a paper is published, there have usually been many rounds of revisions and feedback and the hope is that any serious flaws in methods or data analysis will get noticed and corrected. 

Hoogerwerf: 

Right, so the editor of this journal decided to send a copy of the paper over to Larry, just as a kind of professional courtesy. 

Molnar: 

I would not have been an appropriate referee in the sense of being not a neutral observer there. But he was thoughtful enough to send me a copy and just on the side asked me, “Well, what do you think of this?”

Stults:

And he looks. And at first he really can’t figure out what’s going on with the 1999 data point. It seemed to be the same data set. And so he re-runs his numbers, but still gets the same thing he’d gotten the first time.

Hoogerwerf: 

Still conflicting with the new paper.

Stults:

Yeah. And so he writes back to the editor. 

Molnar: 

I said, “I don’t understand what’s going on here. Did they really use this reference? Because this reference gives this answer.” And he sent back to them and said, “Did you really use this reference?” And they actually then sent back and said, “well, strictly speaking, no. We used a preprint.”

Hoogerwerf: 

A preprint?

Stults:

Yeah, so sometimes scientists are really eager to get their work out there and since peer review takes a long time, they sometimes release what’s called a preprint, which is kind of like a rough draft of the paper before it’s undergone peer review. 

Molnar: 

And so when I heard ah, the second version of this paper, and then I looked online and found a copy of that preprint and compared the two and then I could see that the two papers had two different numbers in them.

Stults:

So then he understands how they could both be doing the right analysis and still come up with different results.

Hoogerwerf: 

So who’s number is right? It seems like the final published version would be the one to use. 

Stults:

It would seem that way. And so Larry, being the ever thorough scientist that he is, goes and looks at both data sets. And he realizes that the timestamp for this one datapoint is different in the two versions of the paper. The two values are suspiciously but exactly off by 12 hours or half of a day. So one of these numbers must simply be a mistake, human error. At this point it was easy for him to figure out which number was right. In half a day, stars move a lot in the sky and he could easily go and see exactly where the stars were in the sky during that 1999 observation.

Molnar: 

And I realized then that I had proven myself wrong. The data that I used was based on a number that would have had the star below the horizon. They certainly didn’t take data that way. The data the other folks used had it above the horizon. So I actually wrote back to the editor once more, and I said, “I think they’re right. And they need to add a paragraph to prove that they are right. Because there are these two versions of the paper. And if they make the argument that I had just done, it will prove, in fact, that they are right, and that I am wrong.” And they ended up adding that paragraph to the paper and acknowledging me at the end of the paper saying, “the critical paragraph was suggested to us by Professor Larry Molnar, we thank him.” 

Hoogerwerf: 

So that’s it. No explosion? 

Stults:

No explosion. That data point being in a different place means that Larry’s entire model and the prediction that came from it just don’t work anymore.

Hoogerwerf: 

So many years of work derailed by what essentially amounts to a typo. So was this whole project a failure?

Stults:

I mean, maybe? In some ways?

Molnar: 

So was it a failure? It was the best we could do given the vacuum of information at the time. And it was the occasion for us to really put our minds to it to come up with a new idea, and really understand how these work. And we have come to that understanding in the time since the failure of the prediction. So the prediction was a failure. But the progress has been quite exciting.

Hoogerwerf: 

Let’s go back to our dots on a chalkboard analogy. So Larry had all these dots and they all fit nicely onto the circle model. But then comes along this other group and says, ”Hey, actually this dot that you have on the circle, that’s wrong, it’s actually out here a little bit.” And they’re right. All of sudden this model that seemed powerful doesn’t work anymore. 

Stults:

So the failure was in the model. But the goal of science is not to prove that your model is right, it’s to have a model that works. So all that means is that they have to go back to the drawing board and start finding a new model that does fit the data. Which inevitably involves getting even more data. So that’s what they’re doing now. 

Molnar: 

Well, having ruled out all the other possibilities, they remained ruled out. Our new possibility is also ruled out.

Hoogerwerf: 

That’s the one that had the stars exploding in a luminous red nova.

Molnar: 

So we need to come up with something else. 

Stults:

And we could end the story there. But trying to end the story anywhere is going to be problematic. If we’re looking for a happy ending, we might need to wait a little longer. 

Hoogerwerf: 

We tend to like stories that end with confetti, or whatever the analogous celebration in the astronomy field looks like. 

Stults:

I mean, an explosion would have been pretty fun.  

Hoogerwerf: 

Yeah I guess you’re right. 

Stults:

But we could shift our expectations a bit and say that a happy ending is that the scientific process ruled out a bad model. And now we’re back on track.

Hoogerwerf: 

But this brings up a bigger question, which is about when we can trust science and the claims scientists make. I mean Larry had this model that seemed really powerful, enough to make a big claim, maybe even risk a bit of his reputation on. And a typo made it all fall apart. So I can see the temptation, especially for non-scientists, to feel like at one point someone will find an error somewhere and everything we know about climate change will fall apart, or all of evolutionary theory will crumble. 

Stults:

The answer to that question is complex. I mean, for one thing you can’t really compare the research on one star by one person to a theory that has been developed by thousands of scientists over decades or even centuries. It’s just a totally different scale. But it might be also helpful, when trying to understand why to trust science as a whole, to understand how science deals with big new ideas. And for that we’re gonna need a philosopher.

Hoogerwerf: 

Is Jim back from sabbatical?

Stults:

No, not yet, so we’re going to have to bring in a guest philosopher. His name is Thomas Kuhn.

Hoogerwerf: 

Ah, Thomas Kuhn. He’s no Jim Stump but I guess one of the most influential philosophers of science of the twentieth century will do in a pinch. I read his really famous book in grad school. 

Stults:

The Structure of Scientific Revolutions.

Hoogerwerf: 

Which is one of the most cited academic books of all time.

Stults:

So Kuhn put forth this idea that science has periods of, what he called “normal science” and other periods of revolution. In the periods of normal science, it’s kind of like all the rules stay the same, all the tools that scientists use are similar, and there’s a certain set of shared assumptions scientists are working with about how the world works. 

Hoogerwerf: 

Kuhn called this a “disciplinary matrix”. A matrix is like a structure or an environment, think like the frame of a building with supports and cross beams and electrical wiring. 

Stults:

And scientists all work within that matrix, trying to solve all the little problems. But in that period there will inevitably be some problems that can’t seem to be solved using those rules and tools. 

Hoogerwerf: 

Occasionally science undergoes a revolution where a bunch of the rules change, a new matrix is developed, in ways that allow us to solve some of these bigger problems. So if you want to build a huge addition onto the top of your house you might need to redo all the supports, rebuild a bunch of the matrix. That’s a “paradigm shift,” Kuhn’s famous term.

Stults:

And this way of thinking about science can be really helpful, as is implied by the fact that Kuhn’s book is cited so often. Though we should say that not everyone agrees with Kuhn. 

Hoogerwerf: 

I don’t think any philosopher has ever said anything that everyone agreed with. 

Stults:

Good point. All that to say, Kuhn’s arguments have a layer of complexity that we probably can’t get fully into here, but we think that at least this idea of normal science and paradigm shifts can be a helpful way of thinking about how science works. 

Hoogerwerf:

So how does this work in practice? 

Molnar: 

Le Verrier actually is a very interesting example of that.

Stults:

That would be Urbain Jean Joseph Le Verrier. 

Hoogerwerf: 

He sounds French. 

Stults:

Good assumption. He was an astronomer doing work in the mid 1800’s. Uranus had recently been discovered, but its orbit was a little different than astronomers expected based on Newton’s description of gravity. Le Verrier hypothesized that there was another planet beyond Uranus that was gravitationally tugging on it and he predicted exactly where that planet was in the sky. And just like that, he discovered Neptune!

Molnar: 

And as a young man, he did this steeped in Newtonian physics, which was new on the scene in which as a young man, he wanted to understand well, and apply and he did successfully. He sent his prediction to the folks in Berlin, who had a telescope who could test it, and they discovered a new planet, based on his observation: what a huge success. I mean, that makes your whole career. But then, he spent the rest of his life trying to repeat that.

Stults:

So the matrix, to use Kuhn’s language again, is Newtonian physics. And it turns out that those rules worked well enough for Le Verrier to predict the existence and position of Neptune. 

Molnar: 

But there was another planet at the other end of the Solar System, Mercury, whose orbit wasn’t quite right: it was precessing just a little bit too fast. So from his view, either Newton is wrong—can’t be—or there must be another planet. Now it’s gonna be a hard one to find because it will be so close to the sun. Let’s call it Vulcan. That’s what they called it. And we will have to look real hard to find it. And he spent the rest of his career trying to prove that precession was Newtonian precession, never finding Vulcan because, well, Vulcan isn’t there. And it was up to Einstein to say no, the precession means Newton really is wrong. And General Relativity replaces Newtonian mechanics and gives us that little bit of extra precession. So with Uranus, he was lucky to say, “Let’s ignore the discrepancy because Newton is right.” With Mercury, he was wrong by ignoring the discrepancy he could have foreshadowed Einstein if he had given in earlier.

Hoogerwerf: 

So Einstein comes along and changes the whole game. He disrupts the previous matrix with his ideas. And what he comes up with is a more powerful model for gravity.

Stults:

And as Larry points out, Le Verrier was right at the cusp of that. The scientific matrix of Newtonian physics worked to find Neptune, but what we know now is that it was never going to work to find Vulcan. He needed a new, stronger matrix, which Einstein would later provide. But he was stuck with the old way of thinking and could never quite get past it.

Molnar: 

To me, I think—and it’s sort of reflections on Kuhn that make me think this as well—is that it is difficult to be as open minded as we should be. And that it is often the case that it takes a generation to be truly open minded on a big question.

Hoogerwerf: 

That’s a good story, but why should it make me trust science any more? And what does it have to do with Larry’s stars?

Stults:

Well if we bring this back to models like evolutionary theory or climate change, we can put this in context and see what happens when new ideas come along in science. And part of what we see is that we don’t have to restart from scratch every time there’s a paradigm shift. Even before Einstein came along with his theory of General Relativity, the prior model, in this case Newtonian physics, was really powerful in explaining how things worked in a lot of ways. Einstein saw that it wasn’t complete, and relativity really did change how we understand physics. But that doesn’t mean that every claim based on Newtonian physics was suddenly wrong. Actually Newtonian physics still works. It really is all we need to build bridges and buildings and for doing the vast majority of everyday, earth-based physics stuff. We still use it all the time. 

Hoogerwerf: 

So we can trust scientific models because even when big changes come along in how we explain things, it doesn’t mean the stuff we knew before inevitably falls apart. Bridges don’t collapse when we realize that time and space aren’t constant across the universe. 

Stults:

And when two stars don’t end up colliding as predicted, we don’t have to worry about other things science has figured out like that life on earth evolved from a common ancestor or that fossil fuel emissions are leading to a warmer climate, suddenly being wrong. Larry made a claim about some very specific stars. He had a good amount of evidence to back up his claim. He put that evidence out there, and the scientific community responded. 

Hoogerwerf: 

Sometimes it can be hard, especially as a non-scientist reading about scientific claims in Facebook posts, to know what kind of evidence backs up a claim. And that’s where it’s important to understand how this process works, to realize that a healthy amount of skepticism is always important when it comes to hearing about some scientific claim, especially very specific claims. 

Stults:

And we’re not all equipped to go look at the journals and critique all the methods for every little scientific thing we hear, but that’s where we need to remember that the scientific process is still dependable because it is done by a community of people challenging each other, testing, retesting and refining ideas. We begin to build increasingly good models of the way things work. One way I like to think of it is that we can trust the process of science (human errors aside) even when the specific claims of science occasionally turn out to be wrong.

Hoogerwerf: 

So there’s another question that listeners might be asking by now, which is what does any of this have to do with faith? 

Stults:

Well knowing how science actually works can be really helpful when putting these two ways of understanding the world next to each other. Faith and science are two important ways of coming to understand the world we live in. But they come to truth in really different ways and it’s important to understand some of those differences and to know where they overlap and where they don’t. With science, what we’re after is the model that best explains the way things work. Our trust in God, on the other hand, is based on a relationship and on a commitment to living within the story of the gospel. So our scientific understanding might change based on new evidence and might have us rework major scientific ideas. For example, when the evidence became overwhelming that the Earth was not in the center of the universe, we had to let it go.  

Hoogerwerf: 

And that’s not easy to do. 

Stults:

Definitely not. Even with things smaller than the structure of the universe. Le Verrier could never get there with Vulcan. 

But Our trust in God and our commitment to the truth of scripture doesn’t change in that way. No scientific model is ever going to tell us we need to let go of our trust in God or our understanding of who God is. In fact, it just can’t really speak to the issue at all. It’s just outside of the domain of science. Science is a powerful tool, but there are many questions that it is just not suitable to answer. 

Hoogerwerf:

Like why do we exist? What’s our purpose? 

Stults:

Or does God exist? 

Hoogerwerf: 

There might be times when these two ways of knowing overlap. We might learn something from science that asks us to go back and rethink our interpretations of some Bible passages, like it did when we learned that the earth orbited around the sun instead of the other way around.

Stults:

Similarly our faith commitments might motivate us to ask certain questions about the way the world works, things that are testable.

Hoogerwerf: 

Like what is happening in our brain when we pray, or how do you heal a particular disease, or what effect our actions have on other creatures and the planet. 

Faith:

And those answers might circle back around, asking us to behave in certain ways because of what Christ asks us to do in the world: to heal and serve and love. 

Hoogerwerf: 

And for Larry, his faith is a big part of why he does science.

Molnar: 

As a Christian, it seems to me what’s important is that my worth doesn’t come from my ideas, my worth comes from what my savior has done for me. So I can have an idea. And it can be wrong. And it doesn’t make me less of a scientist, it doesn’t make me less of a person. Because it wasn’t that idea that made me who I was. And so I think Christians, if they understand it right, are particularly well suited to science. Because we have a stake in it. And it’s the universe that we live in. It’s the universe that God created. But we don’t have a stake in our ideas. That way, we have that openness to discover how is it done? How did God do it? And that’s exactly where we need to be if we’re going to make that progress and not get stuck.

Stults:

Where we left off with the story about Larry’s star was that his prediction for a red nova had failed and yet, the cause of the changes in the orbital period for that contact binary system were still not explained. 

Molnar: 

So we need to come up with something else. And we actually have a thought for that. 

Hoogerwerf:  

Ah good. So we’re not done yet. 

Stults:

Not by a long shot. Though we’re coming to the end of the story we can actually tell, because the rest will be told as time goes on. But here’s the idea:

Molnar: 

The essence of it is that it is related to a magnetic cycle. The presence or the absence of that magnetic field, can cause a small change in the orbital period through something known as the quadrupole moment, something I don’t really want to discuss in detail here. 

Stults:

And Larry has now found even more data, back all the way to almost the year 1900. And it’s starting to form a new hypothesis, and a new model is emerging that looks like it could explain the changes in the period for this star system all based around magnetic fields. 

Molnar: 

The earlier data was not nearly good enough to show the magnetic cycle. But the more recent data is. And so if we follow it over the next change or two, which say the next five or 10 years, we should be able to actually have solid evidence for, or against, this new idea.

Stults:

Without going through the long process of ruling out all the possibilities, including the failed prediction, Larry wouldn’t be where he is now, with new information and on the brink of a new discovery.

Molnar: 

Not as dramatic as knowing how these stars ultimately die and blow up. Which is still sort of our first project. But yeah, worthwhile and worthwhile because it might also be working on other stars. It’s just a hard thing to show for sure. And so you will need a lot of data from a small telescope.

Hoogerwerf: 

Which happens to be something that Larry has access to. 

Stults:

Stay tuned.

Credits:

Stults:

Thanks to Larry for telling his story. 

Hoogerwerf: 

Language of God is produced by BioLogos. It has been funded in part by the Fetzer Institute, the John Templeton Foundation, and by individual donors who contribute to BioLogos. Language of God is produced and mixed by Colin Hoogerwerf. That’s me. Nate Mulder is our assistant producer. Our theme song is by Breakmaster Cylinder. 

BioLogos offices are located in Grand Rapids, Michigan in the Grand River watershed. If you have questions or want to join in a conversation about this episode find a link in the show notes for the BioLogos forum or visit our website, biologos.org, where you will find articles, videos and other resources on faith and science. Thanks for listening. 


Why Trust Science

For centuries, science has increased our understanding of the world around us, yet for some, science feels untrustworthy. But we can discern when science is credible by understanding when scientists are working within their field of expertise, checking their work with others, and whether a particular finding builds upon the confidence that comes from a multitude of data and a community of those searching for truth.

 

Featured guest

Larry Molnar

Larry Molnar

Larry Molnar is a professor in the Department of Physics and Astronomy at Calvin University. He earned his PhD from Harvard University and completed a post-doctorate at the Harvard Smithsonian Center for Astrophysics. He has published many papers and was the subject of the documentary Luminous, about a prediction of a red nova that he and a group of students made in 2017.


16 posts about this topic

Join the conversation on the BioLogos forum

At BioLogos, “gracious dialogue” means demonstrating the grace of Christ as we dialogue together about the tough issues of science and faith.

Join the Conversation