In search of an arsonist: On testing for intelligent design

Posted on 11/05/11 21 Comments

Ryan asks: “What sort of experiment would you run to test for intelligent design?”

That’s a nice, simple question. And to answer it, we need not draw upon any high falutin’ concepts. We can keep things nice and simple.

Consider the devastating May 2011 fire in Slave Lake, Alberta. It was a fire that burned nearly 1/3 of a town of 7000, leading to hundreds of millions of dollars in damage. Immediately after the fire was put out the investigators went in. Just this week they released a shocking conclusion: the fire at Slave Lake was the result of arson.

Now you might ask yourself: how did the investigators conclude that the fire was the product of arson? The answer, of course, is really quite simple: they eliminated all the other explanations.

And there you have it: a straightforward answer to your question of how one tests for intelligent design. One does so by eliminating all non-intelligent causal explanations so that the only plausible causal explanation that remains is an intelligent one.

Now this brings us to the depth of irony of those who critique intelligent design. You know what they commonly say? “Appealing to intelligence is a science stopper!” Oh really? Try telling that to the investigators of the Slave Lake fire. Do you think that once they’ve identified intelligence as the cause of the fire that they are going to call off their investigation? On the contrary, it is at this point that the investigation kicks into high gear as the RCMP (Canada’s national police service) are called in. Having concluded that the fire is the result of intelligence, the investigators are now even more interested to understand who started the fire and how.

In closing, note one more point: it is still possible that the fire could turn out to be the result of non-intelligent causes. Perhaps it is not arson after all. Perhaps the investigators missed something and that will come to light in the days to come. So we recognize that for now the inference to an intelligent cause is a provisional one. But we also recognize that in light of all the data it is the best causal explanation available.

And that’s how it is for intelligent design inferences generally. If you’ve exhausted known non-intelligent causes then the best provisional judgment is that the cause was an intelligent one. It may turn out that the cause was non-intelligent after all, but that doesn’t change the fact that at present the best explanation is an intelligent one. And once we make that inference to intelligence we can expect the investigation not to stall but rather to intensify.

Now how can it be that some people think this logic, which is wholly legitimate in the investigation of a fire, is somehow illegitimate when it comes to the investigation of a cellular structure or the origin of the universe? To be frank, I have never heard a plausible explanation for how and where this line is drawn.

 

Share
  • Katie

    “One does so by eliminating all non-intelligent causal explanations so that the only plausible causal explanation that remains is an intelligent one.”

    That may be a valid way to demonstrate the truth of something, but it’s not a test in the scientific sense of the word. One can build the case for arson by providing both positive and negative evidence; the evidence for ID is entirely negative. In science a hypothesis is tested by attempting to nullify it, which is generally done by trying to demonstrate the failure of positive evidence to uphold the hypothesis. So when the hypothesis can only be inferred from the failure of other hypotheses, it can’t really be tested per se. The only theoretical way to prove its truth would be to demonstrate that it is the only remaining possible explanation, which is not really possible in the case of origins questions.

    This is highly problematic from a scientific point of view. However, as I believe you’ve pointed out, ID is not really a question of science but of philosophy of science. Scientific testability is not the criterion for truth; it’s just a way of making it easy to test whether a belief is true. ID may very well be true in some cases, but it’s going to be impossible to demonstrate that in the typical scientific manner.

    • randal

      “That may be a valid way to demonstrate the truth of something, but it’s not a test in the scientific sense of the word.”

      “Test” within the realm of science has multiple senses. Indeed, defining and deliminting the boundaries of what is, and is not, science is one of the most controversial and hotly debated areas in all of contemporary philosophy.

      “So when the hypothesis can only be inferred from the failure of other hypotheses, it can’t really be tested per se.”

      Of course it can. Take Behe’s concept of irreducible complexity. We test the hypothesis that a given structure is irreducibly complex by attempting to falsify it via the demonstration of the reducible complexity of the structure in question.

      “However, as I believe you’ve pointed out, ID is not really a question of science but of philosophy of science.”

      Correct. And if we accept ID then it means that inferences to intelligence are explanatorily legitimate in scientific enquiry.

      “Scientific testability is not the criterion for truth….”

      Right you are. So while we can continue to debate whether or not ID is “science” (and who defines what science is), the even more important question is whether it is true, i.e. whether there are parts of the natural world which are correctly explained by appeal to intelligence.

      • Katie

        “So while we can continue to debate whether or not ID is “science” (and who defines what science is), the even more important question is whether it is true”

        Agreed.

        The reason I disputed your definition of testing is because I think it becomes a distraction. Materialists can pretty easily dismiss ID by construing the IDers as trying to redefine the scientific process to include things that science can’t speak to. What we really need to discuss is whether ID is plausible, not whether it falls under the standard definition of scientifically testable. I’d rather see ID proponents concede that the theory is not scientifically testable in the same sense that cell theory (or arson) is, and then move on to the more fruitful discussion of whether or not ID seems to be the most reasonable explanation of the data.

        • pete

          But wouldn’t irreducible complexity be falsifiable?

          Couldn’t scientists take the supposed object and see if it could be further reduced?

          I may be wrong, but that seems falisfiable to me.

          • http://ingles.homeunix.net/ Ray Ingles

            Yup, irreducible complexity is falsifiable, and all the putative examples put forth for it have been falsified.

            • randal

              Wow, you’re familiar with every proposed example of irreducible complexity? You must really keep on top of the literature! You wouldn’t happen to have journal references would you?

  • http://www.atheistmissionary.com/ The Atheist Missionary

    If you’ve exhausted known non-intelligent causes then the best provisional judgment is that the cause was an intelligent one.”

    No sir. The best provisional judgment is to look for (an as yet) unknown non-intelligent causes while, in fairness, not dogmatically ruling out an intelligent one. Why? Because, to date, we have no evidence to suggest that intelligence is the cause of any existent piece of matter aside from human artifacts. As you well know, Dembski has likened the scientific search for ID in biological systems to SETI and cryptography. I am fine with that. I hope ID theorists move from theory to experimental reality and show me the money.

    • randal

      I’ll respond to this in a blog post.

  • Jag Levak

    “… how did the investigators conclude that the fire was the product of arson? The answer, of course, is really quite simple: they eliminated all the other explanations.”

    A scientific arson investigator (and the field is sadly rife with the unscientific sort) would first try to determine the cause. Failing to find a cause would simultaneously mean finding no natural/accidental cause and finding no deliberate cause, and the case would simply be logged as undetermined cause–an entirely legitimate finding in arson investigation. It would be the worst sort of arson investigator who can’t find any cause at all, and thereby concludes it must have been arson.

    “And there you have it: a straightforward answer to your question of how one tests for intelligent design. One does so by eliminating all non-intelligent causal explanations so that the only plausible causal explanation that remains is an intelligent one.”

    For an example of how quickly this can go wrong, consider the case of seemingly spontaneous atomic decay. In adding up all the energies that emerge from such a decay event, we find no surplus, and there is nothing in atomic theory which suggests an internal cause, so for the time being, we are logging this phenomenon as having no known cause. What the principle above suggests is that we should say we have eliminated all non-intelligent causal explanations, so intelligence remains as the only plausible causal explanation, which is just dumb. In the first place, we can only eliminate known causes. There always remains the possibility of unknown causes. In the second place, even if it were possible to eliminate all causes, that would eliminate all intelligent causes right along with all the non-intelligent causes.

    “Now this brings us to the depth of irony of those who critique intelligent design. You know what they commonly say? “Appealing to intelligence is a science stopper!” Oh really? Try telling that to the investigators of the Slave Lake fire. Do you think that once they’ve identified intelligence as the cause of the fire that they are going to call off their investigation? On the contrary, it is at this point that the investigation kicks into high gear as the RCMP (Canada’s national police service) are called in.”

    Which strongly indicates they found a cause for the fire, and the cause has properties common to deliberately set fires. Because if all they truly have is the lack of a finding of a cause, that would leave the RCMP with nothing to investigate.

    “In closing, note one more point: it is still possible that the fire could turn out to be the result of non-intelligent causes. Perhaps it is not arson after all.”

    This is a serious risk when dealing with an animal which has a propensity to see agency, even when it isn’t there. (see the Cameron Todd Willingham case for a particularly egregious example of how far things can go wrong when an investigation is conducted without sufficient scientific rigor.)

    “So we recognize that for now the inference to an intelligent cause is a provisional one. But we also recognize that in light of all the data it is the best causal explanation available.”

    In the case of arson, we are usually dealing with probable causes. Deliberately set fires are included among the candidate possibilities because we know they happen, they happen a lot, and humans have several common ways of setting fires which have characteristics readily distinguishable from natural causes. When you find chemical trace and burn patterns consistent with gasoline being splashed on the walls of a home, and the burned remains of a gasoline can in the living room, we don’t conclude that was a deliberately set fire on the basis of it being utterly impossible for natural causes to have left that evidence (since we don’t know that), but because the arson explanation is the most probable and least remarkable of any of the more common explanations. We don’t have an analog to that in biology.

    But there is another sort of error which can occur in arson investigation. A skilled arsonist can set a fire in a manner which leaves evidence that is entirely consistent with natural or accidental causes, so there is always a possibility that what looks unintelligent is actually intelligence intelligently concealing itself. And if humans have the ability to conceal their tracks from humans, how much moreso could a cosmic intelligence?

    “once we make that inference to intelligence we can expect the investigation not to stall but rather to intensify.”

    If ID proponents would care to lay out exactly how such an investigation could proceed, maybe they would actually get somewhere in the scientific community.

    • randal

      “It would be the worst sort of arson investigator who can’t find any cause at all, and thereby concludes it must have been arson.”

      Spoken from the expert on arson investigation. It isn’t that they didn’t find any cause at all. It is that they eliminated all known natural causes until the only remaining known cause left was arson. That’s how they found the cause: through process of elimination. As I said, it is possible that this could ultimately be mistaken: maybe they missed something. But that doesn’t change the fact that at present they have adequate evidence to draw a design inference.

      “If ID proponents would care to lay out exactly how such an investigation could proceed….”

      Obviously you haven’t read the proponents of ID. Start with Meyer’s book since he addresses this at length.

      • Jag Levak

        “It isn’t that they didn’t find any cause at all. It is that they eliminated all known natural causes until the only remaining known cause left was arson. That’s how they found the cause: through process of elimination.”

        Even if the investigator may have characterized it that way to the press, I really doubt that was the procedure. I expect the first thing they did was backtrack the fire damage to find the point of ignition. Next would have been to inspect the area around that point for clues as to what could have started the fire. I’m pretty sure at this stage they would have logged all evidence which could possibly pertain to ignition, rather than focusing on natural causes exclusively to eliminate them first. They would look for cigarette butts, signs of a campfire, footprints, vehicle tracks, anything which can reflect or refract light (a plastic bottle of water can start a fire, and then be consumed in the fire), proximity to power lines, proximity to coal seams, volcanism, or other underground heat sources, compost piles, indications of lightning strike, rock falls, bullets, machinery, welding sites, discarded chemicals, etc. etc.

        If they find no cause at all, that should properly be logged as undetermined cause. To have a finding of arson, they first need to find that ignition was due to some human activity–which would be established by evidence of human activity. Then they would need to establish that the activity had the deliberate intent of starting the fire, which means looking at the characteristics of the evidence, not looking at a lack of evidence for some other sort of cause. Finding no natural cause, but finding lots of indications of human activity still isn’t enough to establish arson. Most places segregate fires due to human activity into accident (not reasonably foreseeable), negligence (carelessness in the face of a reasonably foreseeable risk) and arson (setting fire with deliberate intent) and again, these determinations are made based on the characteristics of the evidence. Unless Canada is really weird in this respect, eliminating all natural causes shouldn’t be nearly enough to establish arson.

        • pete

          So using the arson investigation, from experience myself, its a bit of science involved, and analogous to ID.

          Jag is right that not all natural causes are prima facie ruled out. If propellant used for ignition is found in a manner consistent with arson, it is at that point that criminal causes are sought to be ruled out.

          So it isn’t a jerk move for a person to reason that if the universe appears to be arranged in a manner that is consistent with an intelligent cause, that it cannot be infered.

          However, the faithful truth seeker will be open to investigating non-intelligent causes for what appears to be intelligent.

          A better analogy would be a case of a murder. Because the stakes are so high, the cop wants to be sure that they don’t arrest an innocent person.

          So they seek all avenues of investigation. But even if they don’t have a DVR with sound recording replay of the incident, if the evidence supports, beyond a reasonable doubt, the case that an intelligent cause killed MR.X, than they go with the arrest and charge.

          Our job, as personal judges, boils down to how we filter, weigh, and discern all evidences.

  • http://ingles.homeunix.net/ Ray Ingles

    I don’t know if I’d jump to using arson investigations as a model. It’s not exactly clear that it rises to the level of ‘science’ quite yet.

  • Ryan

    Randal,

    Wow! I only posted two times and already I got my named mentioned in your blog :D

    While I appreciate your taking the time to try and address my question I was hoping for an example of a test for ID and not an analogy. Can you think of one?.

    I don’t think your analogy works because we can collect evidence and study natural fires and purposeful acts of arson. We know that people set fires deliberately, so we’re not invoking an unknown which wouldn’t be the case with ID since we have never witnessed the creation of life.

    The determination that a fire is deliberate isn’t just about ruling out natural causes but also identifying intentional causes

    http://www.theglobeandmail.com/news/national/700-million-slave-lake-fire-caused-by-arsonist/article2221179/

    “On Tuesday, the province said the fire was deliberately set and that investigators have pinpointed the ignition site and cause – though they wouldn’t reveal either.”

    These are some of the problems I’m wrestling with when it comes to ID. I can’t think of a way for ID not to be a GODDIT (God of the Gaps) type answer.

    I’m hoping you can describe or identify a test that will give me a good answer.

    • randal

      “While I appreciate your taking the time to try and address my question I was hoping for an example of a test for ID and not an analogy. Can you think of one?”

      This wasn’t merely an “analogy”. It was one example of how people test for ID in the natural world. The point is that testing for intelligence is something that we do all the time. It is not some novel, special pleading on the part of the advocate of ID. In fact, we’re all advocates of ID. The question is whether there is any justification for denying the licitness of inferring to intelligence in certain areas of inquiry.

      You purport to offer a disanalogy which you believe to be relevant: “We know that people set fires deliberately, so we’re not invoking an unknown which wouldn’t be the case with ID since we have never witnessed the creation of life.” If I understand you correctly, your point is that it is illegitimate to infer intelligence when we’re not familiar with the source of the intelligence. But that is obviously false as I demonstrated with my irrepressible wit here: http://randalrauser.com/2011/07/intelligent-design-unknown-intelligence-and-a-ouija-board/

      If you want to see a case for ID applied to a particular issue you should take the time to read Stephen Meyer’s Signature in the Cell where he argues for intelligence as the best explanation for the highly specified information in DNA. It is, as he argues at length, the only known cause that can produce this kind of effect. This doesn’t preclude the possibility in principle that some other cause could be discovered in the future. But of course the same could be said for any scientific question. The point remains however that at present the only known cause of the perceived effect is intelligence.

      Finally, you worry about God of the gaps. I’ve written about that here: http://harmonist.us/2009/10/whats-wrong-with-god-of-the-gaps-theology/

      • Jag Levak

        “It was one example of how people test for ID in the natural world.”

        That didn’t look like any test to me. What I saw was a guess made in the absence of clear support for some natural cause (which, by the way, looks a lot like a God of the gaps approach). At most, that would make it a hypothesis. But to test a hypothesis, you first need to work out the necessary implications, and from the implications you make predictions about what we should see and what we should not see if the hypothesis is true, and then you go out and compare the predictions to actual findings. Ideally, the predictions should be clear and distinctive. Clear so that we can tell when a prediction is successful or fails, and distinctive so that successful predictions don’t simultaneously support a host of other competing hypotheses making the same prediction. With no critical failed predictions and reasonably consistent and repeatable findings in support of distinctive predictions, we would have a good performing hypothesis with a real scientific basis.

        Now, ID arguments that says naturalism alone is insufficient (God of the gaps approach) or that say what we see has some things in common with artifice or known products of intelligence (argument from analogy) may be well and good for lending support to the plausibility of the ID hypothesis, but if that’s as far as it goes, then all you’ve got is a notion that more properly belongs in philosophy. It’s got no place in science until it can be tested scientifically.

        • randal

          “ID arguments that says naturalism alone is insufficient”

          How do you define “naturalism”?

          • Jag Levak

            [“ID arguments that say naturalism alone is insufficient”]
            How do you define “naturalism”?

            These would be ordinary physical processes of the sort normally studied by science. In your fire analogy, it could be ignition causes like lightning, rock fall, composting, meteorite, and volcanism. Examples in ID that use this approach are irreducible complexity (systems which supposedly cannot be assembled piecemeal by evolution), and claims that there are no material causes which have the ability to produce large amounts of “specified information”.

            I’m aware you sometimes argue that “natural” does not preclude “intelligent”, but that position poses a problem for your fire analogy, since ruling out non-intelligent causes would depend on the ability to identify sources as not being the product of intelligence, and it isn’t clear how that would work in a universe where everything is the product of intelligence.

            • randal

              I’m referring here to methodological naturalism. As I’ve noted, that is a difficult concept. Sometimes it is used to exclude intelligence of any type. Other times it is used to exclude appeal to “supernatural” intelligence. But then that begs the question of how to define “supernatural” intelligence which compounds the difficulties for methodological naturalism and suggests that they are better off dropping those arbitrary, vague and question-begging stipulations.

  • http://www.thepietythatliesbetween.blogspot.com Eric Reitan

    In case anyone is still interested in this post a month later, I just wrote up, on my own blog, a rather lengthy reflection/response on the proposal offered here–to wit, that we test for intelligent design by ruling out all other plausible alternatives. It can be found here.

    • randal

      Thanks for the heads up. I’m very interested.