For a while, there was an attempt to push "Intelligent Design" as a purely scientific (and not at all religious) alternative theory to evolution through natural selection. That effort seems to have declined after the ruling in Kitzmiller v. Dover Area School District, which found ID to be a transparent attempt to disguise religious teachings as Sciantificness. Which, of course, it was, but could there be a genuinely scientific inquiry about the concept of Intelligent Design? I'd like to suggest one approach that could work.
To begin with, we need to define what we mean by "intelligent design". I think it's fair to say that the concept is derived from analogy to human intelligence and purpose. In the classic example, when we see an object like a pocket watch, that seems to have been put together to perform a very specific function, we infer that some deliberate action -- by a watchmaker -- explains why these bits of matter happen to be arranged in just the right way to perform that function. Therefore, goes the argument, when we see matter organized into eyes to perform highly complex functions like vision, it makes sense to infer an eye-maker. (I have a lot to say to criticize the assumptions of this argument, but I'm not going to do it in this post, because I want to talk instead about how we might try to apply it as an actual scientific theory. I may eventually do so in a subsequent post.)
Now, the mere existence of objects with apparent purpose in the universe like eyes and legs and DNA does not necessarily show that ID is right, because after all, Darwin's theory also explains that. (Indeed, if there weren't such amazing complexity and highly-tuned function in the biological world, there'd be nothing for Darwin to explain.) What we need is for something that ID predicts that Darwin doesn't.
I spend quite a lot of time thinking about intelligence, and my wife is a speech-language pathologist who works with young children and their language development, so we often find ourselves talking about how minds work and learn, and yesterday she was telling me about a little boy who she had manage to help memorize the alphabet a couple of months ago, and who now is actually reading. She could tell he was actually trying to decode the letters on the page, instead of simply reciting the nursery rhymes from memory, because he had started to make the right kinds of mistakes. In other words, she could tell he was making real progress, actually applying his intelligence to the problem, because she could see where that intelligence was making the kind of characteristic errors that only intelligence can make.
So what kinds of mistakes are characteristic of intelligence, as distinct from just random error? This is a very difficult question, but I think we can make a couple of observations about some of them. For example, you may be aware of the horrifying Virgin Cleansing Myth in areas where HIV is common. The idea is that if an infected man has sex with a virgin, he will be cured. The Wikipedia article I linked to there quotes Cati Vawda of the Children's Rights Center in South Africa as saying, "We have no idea where this idea has come from..." but I think I have an explanation, and it has to do with a fairly simple misunderstanding or misparsing of a true statement.
How do you get HIV? Generally by exchanging bodily fluids with an infected person, usually by sex. So, obviously, you can't get it by exchanging bodily fluids with an UNinfected person, right, and (if you assume that you only can get the virus through sex) a virgin is probably going to be uninfected? Well, that's true, but consider very carefully the grammar of how this fact can be expressed in English (and presumably most other languages): "If you have sex with a virgin, you won't get HIV." The sentence (and many other formulations) is literally true on the reading that you won't catch HIV from a virgin, but it is shockingly dangerously false on the reading that sex with virgins protects against infection from other sources.
I don't think the Virgin Cleansing Myth just popped into someone's head and spread like an urban legend. I think that it arises again and again when someone partially hears an explanation of how people get infected, misunderstands it (through a perfectly natural alternate parsing), and shares or reinforces the mistaken understanding among others. It is a mutation, therefore, of a beneficially true existing meme, rather than a wholly new spontaneous meme.
I have seen other characteristically human errors, in my study of chain letters. The Four Reports pyramid chain (Tetralogia vulgaris) is a good example. The basic idea behind this chain is that you order a copy of each of four "business reports"from the four people on the chain letter for $5 each, and then you modify the letter by adding your name and address in position 4, moving everyone else up one position, and removing person in position 1. If all goes as described, you will get $50 from people ordering report #4, and then $500 from people ordering report #3 once your name is moved to position #3, $5000 in orders for report #2 and finally $50000 from people ordering report #1. (By the way, it almost never goes as described. You should consider yourself lucky if you get back the same money you put into it. Minus postage.)
At one time, I had a collection of about 150 unique specimens of this chain letter (lost them in a hard drive crash, alas), and each specimen had its own list of the last four hosts. By comparing the lists of names, I was able to reconstruct the lineage of a majority of my specimens, connecting most of them into a single tree that spanned 10 generations.
But I also an interesting anomaly. While most specimens that included "John Doe" of "123 Main Street, Anytown USA" in any of positions 2 through 4 had "Richard Roe" of "456 Elm Avenue, Blackacre" as the recipient before him, occasionally I would find a specimen that had John Doe in position 2 and someone else (whose name happened to resemble the email address from which I'd received the specimen) in position 1. What I think happened was this: someone misunderstood the way the chain letter is supposed to work, assumed that being in position 1 meant more orders and more money, and decided to put themselves in position 1 sooner in order to make that money faster, not realizing that it just meant they'd be eliminated from the list in the very next generation.
I saw this in four or five instances, and while it's possible they just misread the instructions and mixed up where they were supposed to put their names, I suspect at least a few of them made the specific error I described, given the basic psychology of the kind of person who falls for an easy money scam.
If Intelligent Design is true, that is, if the order in the universe around us is the result of some intelligent being acting with a purpose, one way we might be able to distinguish a designed universe from an evolved one would be if we could find some kind of artifact of the kinds of characteristic errors that intelligent beings typically make. Now, I'm aware of lots of "design flaws" in various living things, but all of the ones I know about are consistent with evolved solutions, so what we'd be looking for is some kind of error that stands out as the kind of error that could only be made by someone, based on some misapprehension about purpose or method, rather than something that can be explained by a Darwinian model.
Of course, any intelligence great enough to create the world around us is going to be far smarter than we are, at least in the relevant ways, so it's going to be hard to find this sort of error. Even so, errors are usually easier to spot in hindsight, after they're made, so it shouldn't be a completely hopeless task. It's only hopeless if we also postulate that this intelligence is perfect and never makes any errors, but that postulate pretty much cuts off any possibility of experimental investigation, doesn't it?