Episode Transcript
[00:00:07] Speaker A: Welcome to ID the Future, a podcast about intelligent design and evolution.
[00:00:14] Speaker B: Greetings, I'm Andrew McDermott. Today I'm welcoming back writer and speaker Kirk Durston. Durston holds a PhD in biophysics, a master's in philosophy, and undergrad degrees in physics and mechanical engineering.
Author of several papers and peer reviewed science and philosophy journals, he has for decades studied the possibility that meaningful information in general, and genetic information specifically, is the fingerprint of intelligence.
One of his current research projects involves working with a team of scientists to develop software that can help us better understand the structure of proteins and how they fold. Kirk, welcome back to the program.
[00:00:53] Speaker C: Thank you, Andrew. I'm glad to be here.
[00:00:56] Speaker B: This is the second of a three part series exploring a written series that you recently offered up at your blog, Kirkdurston.com, that looks at the differences between the major categories of experimental, inferential, and fantastical or science fiction. Today we'll zoom in on your discussion of inferential science, and we'll continue to probe your thoughts on the detectability of intelligent design in nature.
First, just a bit of review for those who may not have listened to part one of our discussion yet. Can you tell us how you divide science into three major categories and maybe why?
[00:01:34] Speaker C: Yeah, I felt the need to divide science into three major categories because I saw a lot of people just totally believing everything they heard that a scientist said. And in real life, I knew that some of those things were excellent and trustworthy. Others things were pretty sketchy, and others were just outright fantasy. And so, as I thought about it, I realized that you can actually divide 21st century science into three categories. One would be experimental science, where you do an experiment, it produces results, you publish those results. Other people then can attempt to reproduce those results. And if they do, then you've got a pretty solid conclusion here. Excellent, excellent results. And that is the gold standard of science. The next aspect of science, next category I call inferential science. And that is where you might start with experiments and observations. But due to the nature of the situation, whatever it is you're investigating, all you can do at that point is then make inferences. Maybe a deductive inference, which is more probabilistic, or an abductive inference, which is an inference to the best explanation, and then come up with a conclusion that you can't actually test because things may be very complex or way too massive for an experiment to take place, or it took place in the past.
So if it's a small inference, if it's highly probable, if it's a good deductive inference, that conclusion might be trustworthy. We don't know for certain, but it's at least trustworthy. But if it's a huge deductive inference, the probabilities are extremely small there, and therefore, we might not want to put our faith in that particular conclusion. And then the third category was simply what I call fantasy science. And that is you have, science comes up with some sort of a model, mathematical or verbal, we can't even test it. And therefore, it's not even in the realm of science. It really falls into the realm of science fiction. And we can write science fiction either using rules of grammar and vocabulary and words, or we can write it using mathematical models or mathematical axioms and techniques, but it still remains science fiction. If we cannot test it.
[00:03:56] Speaker B: We get some great stories out of that category, though, of course, I love.
[00:04:00] Speaker C: Science fiction, so I find them fascinating. I just finished another Sci-Fi book just two days ago. I clearly make the distinction between science and science fiction for sure.
[00:04:11] Speaker B: Well, what options does science have on the table right now to explain the appearance of functional information?
[00:04:17] Speaker C: Well, if we're going to stick with cold, hard science, it may surprise people to say there's only one option out there. And when I say there's only one option out there, what we're talking about is what options do we have to explain the functional information we see encoded in the genomes of life? And to qualify as an option in science, that is, we have to be able to actually observe functional information being produced at a statistically significant level.
And when we make that compl, in other words, it has to be observable, reproducible. Otherwise, it really doesn't belong in the realm of science. And we can come up with ideas.
In science fiction, for example, there can be lots of different ideas, but they're not for interstellar space travel, for example. But they're not observable. They're not testable. At this point, we can actually reproduce that interstellar space drive for that spacecraft that's got to travel around the galaxy. And so that's clearly science fiction. So when we rule out, when we look at what options, what does science, what can we actually observe out there that actually produces statistically significant levels of functional information? There's only one option on the table, and that is intelligence.
And you can produce, and we can test that. For example, I can hand somebody a laptop or a smartphone and ask them to text somebody something, and then we can run that text through a program that will actually quantify the amount of information required for that text. And what we find is that we can produce statistically significant information all the time. Humans can. And it's not because we got thumbs to text with. It's not because we got, you know, a head or it's not even because we got a brain, because lots of animals have brains. Some animals have thumbs. It's none of that. What it is about our brain is that it's highly intelligent. And so if you're going to produce statistically significant levels of information, the only reproducible option science has on the table right now is intelligence.
[00:06:34] Speaker B: Absolutely. And so studying intelligence is how you would study functional information and the appearance of it and even the origin of it. Now, we talk about the scientific method, and often people will talk as if those letters are capitalized, scientific method. The scientific method. But lately, I've come to realize that, you know, the scientific method, as it's traditionally known, works well with experimental science. But there are different methods you'll use with inferential science and in particular, studying historical theories like darwinism and intelligent design. Tell us about the scientific method you use to test for intelligent design.
[00:07:15] Speaker C: Okay, first of all, on the basis of observations that all of us make every day in real life, on that basis, we can come up with a hypothesis. And the hypothesis is that the ability to produce significant levels of functional information is unique to intelligence. Now, the key word there is unique. And when I say unique, that means nothing else can do it. And the moment I say that, we've set ourselves up for possible verification or falsification, because all we have to do is find another way that functional information can be produced that does not require intelligence, and then that hypothesis is falsified. So the hypothesis itself qualifies as a solid scientific hypothesis in the sense that it can be falsified or verified.
Now, the next thing would be to make some predictions to see what follows from that and what are the. There's. It could go into a number of fields in archaeology, biology, forensic science, and so forth. All of it can use that hypothesis.
But in intelligent design, the next thing that I would do is, okay, we have an effect here, and the next thing we need to do is maybe define what we mean by intelligent design. And this is, you know, if you talk to different scientists who. Who are supportive of the idea of intelligent design, you may have slightly different definitions, because there's different ways to approach the problem. But my own definition of intelligent design is as follows. It is an effect that requires intelligence to produce.
And so, for example, of intelligent design, I would point to a smartphone or a supersonic aircraft, or even, in modern medical science, would be medicines that have been produced in a lab by intelligent scientists who were required to be able to figure out that medication and how to make it and produce it. So, virtually everything that modern science has given to us is a result of experimental science. But key behind that was it was intelligent scientists working away.
So once that hypothesis is there, that is, the ability to produce statistically significant levels of functional information is unique to an intelligence. And once we've defined what intelligent design is, and that is, it's an effect that requires intelligence to produce, we then have a method we can use to test for intelligent design or to test that effect. And that is we can quantify the amount of information required to produce that effect, for example, to produce that cell phone or that aircraft or that protein family. And once we've quantified that, we can see, okay, here's a number.
Is this level of functional information significant or not? That is, is this what we would expect just in the normal background processes and noise of nature, or is this an anomaly? Is this something remarkable? Is this something we would not expect to occur with our null hypothesis that there is no intelligence involved and we're just looking at chance and order? And if it does test positive for a statistically significant level of functional formation, then we can conclude, or we can infer that required intelligence to produce, and that's an inference. So, this whole method falls into the area of inferential science, and it's very close to methods used in, let's say, forensic science and archaeology and even Sati, they're using this idea, maybe explained slightly different, but at its heart, it's the same approach.
First of all, measure the effect to see what sort of information is required to produce it. If that effect requires at a significant level of functional information, then it tests positive for having originated or having required intelligence to produce.
[00:11:28] Speaker B: Okay, well, tell us a few things that can go wrong with inferential science.
[00:11:34] Speaker C: Well, I guess, keep in mind that an inference can be classified under different categories. One would be a deductive inference. And a deductive inference is when it's probable, when it has some probability of taking place.
You look at the evidence, and this occurs in forensic science. You look at the evidence, and you see that, oh, Bill's fingerprints are on the doorknob of this cabinet where I was storing huge bags of gold bullion, and that gold bullion is missing. Now, I would infer from that, since Bill's fingerprints are on the knob, that bill was the one who stole the golden. We can't reproduce that, but we can certainly look at the cause. We can't go back in time and actually, okay, just rerun the camera. Now, we can't do that, but we can infer that Bill was the culprit simply because the fingerprints are the fingerprints of bill, and it's highly improbable they belong to someone else. So the probability that bill did it is very high. So that's a good deductive inference. Bad deductive inference is when you start appealing to highly improbable scenarios.
And you can actually see this a lot in, say, evolutionary biology articles. And I have an example here titled the origin of the very first species in the start of darwinian evolution. And in that article, they used 28, what I call lack of data words. These are words you use when you do not have the data to support your inference. So here's some examples. They use the word presumably, probably possible, might have at some time, possible scenario, could have proposed over time, eventually generated, researchers believe, seems likely is conceivable. Now, when you start using words like is conceivable or researchers believe, or eventually generated, what these words are, they're lack of data words. There are words that you use when you do not have the data or the experimental results to back up what you're saying. So in place of that data, you got to use these words. The more of this that goes on in inferential science, the less likely it is or the more improbable their conclusion is. In this case, there's 28 of these words in the space of two pages, 28. So when I see something like that, I'm saying, okay, this is a deductive inference, what they're proposing here. But there are so many things that seem to be quite improbable that I don't really have much rational basis for taking their conclusions seriously. And that's one of the pitfalls with inferential science. And unless people start thinking of, start critically analyzing what sort of inferences are being made and how big of an inferential leap is being made, then they're not in a very good position to evaluate, say, all the popular science articles they see in the media every day. They have to ask themselves, how likely is this? Are they using, and are they using these lack of data words in this article?
[00:14:55] Speaker B: Well, I'm really glad you pointed that out and used an example from an actual article, two pages, and all of those lack of data words, we need to train ourselves to look for those words, identify the inferences being made and ask, are these leaps that are acceptable or unacceptable, given the data? So that's. That's great to practice. Thank you.
[00:15:21] Speaker C: The lights really come on for a person once they begin to identify lack of data words.
[00:15:25] Speaker B: Well, what can we do? And you've touched on this a little bit, but just to wrap up here, what can we do to avoid these pitfalls in inferential science?
[00:15:31] Speaker C: Well, I would say, first of all, become aware of the.
The idea, the category of inferential science. The first question is to ask yourself, are the results hard, actual results that were produced and that can be reproduced by other people, or is the result they're talking about merely a conclusion that they're inferring? Is there a bit of a leap here? Is there some experiments or observations they made and then they inferred, or they took this deductive leap to the conclusion? That's the first thing to do. If there is any kind of deductive leap at all, you're usually talking about inferential science.
The second thing is to safeguard in this, to help navigate through the. What can be a bit of a pitfall in inferential science is say, well, how big is that leap? How probable is it? So, in cases in forensic science, for example, those leaps are probably pretty small. They need to be stand up in court, and so the conclusions might be fairly trustworthy. And, in fact, the court will decide if they're trustworthy beyond reasonable doubt or not. Other areas in archaeology or in SETI or in intelligent design science, now, even with companies such as Monsanto, that design certain types of grain, like canola, for example, and have genetic markers in there, if some person is sowing a variety of canola, which is a type of grain that they make canola oil from for cooking and so forth, and then the company who designed that wants to, let's say, take that press legal action, you then have to take that, the genetics to court, and you have to make the inferences, okay, the markers are here, but how likely is it that this could have arisen naturally, as opposed to the ones that we actually encoded into the canola genome itself? So we're now putting markers in genomes, and this is where inferential science and intelligent design science becomes applicable in biology. When it comes to even in case of legal battles between different people who might have developed this new variety and other people who are maybe plagiarizing it.
[00:17:56] Speaker B: And so forth, that is a compelling example. Yeah. Well, that's all the time we have for today, but we will be back with one more episode where we'll look at the third category that Kirk identifies in modern science, and that is fantasy science or science fiction. And we'll take another look at some of his work on the structure and folding of proteins and what that means for intelligent design. Kirk, thanks for joining us.
[00:18:20] Speaker C: Oh, my pleasure.
[00:18:21] Speaker B: Well, to read more of Kirk's posts, you can go to Kirk Durston Dash s Dash t Dash o, Dash N, Dash N.com. and for more episodes of ID the future, including part one of our discussion, visit idthefuture.com on the web, or subscribe and download through your favorite podcasts app. And by the way, if youre enjoying the content you hear on ID the future, consider giving the show a positive rating on your podcast app of choice. Thatll help us reach more listeners with the compelling evidence and arguments that you hear each week on the show.
Well, until next time, I appreciate you listening. I'm Andrew McDermott for ID the future. Thanks again.
[00:19:00] Speaker A: This program was recorded by Discovery Institute's center for Science and Culture. Id the Future is copyright Discovery Institute. For more information, visit intelligentdesign.org and idthefuture.com.
