Creationists (whether you are one or not) come up with this particular canard repeatedly. Once more.... Science isn't just conducted by looking at the facts and arriving at conclusions. Looking at facts is the first part of the process, and arriving at conclusions is the last. There's a whole slew of stuff imbetween, that creationists tend to ignore.
Indeed. Here's a (simplified) description (and yes, I wrote it myself):
Explaining the scientific method
I've recently noticed that most creationists have absolutely no clue about how the scientific method works, so allow me to give you a *very* simple overview. The reality is far more rigorous than this, and subjected to endless review and retesting, but here's the short version...
Creationists imagine that scientists just gather some data, make up a "reasonable" speculation that they think could account for the data, and then kick back and go on vacation. The creationists think that's all there is to science, just the "speculation" part.
Nope.
That's just the barest *beginning*.
That's called the "hypothesis" stage. Scientists all over the world try to come up with all the hypotheses they can as possible explanations for any given scientific puzzle. This is a good thing -- the more brainstorming, the better the chance that someone will come up with a "speculation" that hits closest to the right answer.
So then they just argue over it to "decide" which one's right, eh? Nope, sorry, that's the creationist version.
Instead, they *test* these various ideas to see which ones actually work when compared against reality, and which ones fail when compared against reality. You've heard of those "experiment" things, right? This is the testing.
But contrary to cartoon-level public impressions of experiments, they're not always done by guys in a lab pouring test tubes together, or by scientists attempting to actually *reproduce* the process they're studying. Those are *one* way to test a hypothesis, but nowhere near the only reliable way. It's not possible to recreate the entire Hawaiian Islands chain in order to test hypotheses about how they formed, for example, but there *are* countless other ways to test hypotheses about geology -- or evolution, or any other field.
That's why you've heard scientists talk about "predictions" so much. One of the most widely useful, and most reliable, methods of testing a hypothesis is to work out its consequences -- to determine what results would occur if that explanation *was* the correct one. If that really *is* how things happen (or happened) in the universe, what "side effects" would it have, aside from the data or phenomenon itself which we crafted the hypothesis to explain? These are its *predictions*.
This is how we test to separate the good explanations from the bad explanations. If the Hawaiian Islands formed as a result of continental drift carrying the Pacific tectonic plate across a crustal magma plume, this would leave many, many kinds of tell-tale results which would be noticeably (and *test-ably*) different than the kinds of things you'd find if some *other* explanation was correct about a different manner in which the Hawaiian Islands had formed.
So to decide between the two (or fifty) potential explanations (hypotheses), you work out the "side effects" (predictions) of each different explanation -- what you'd find if X had actually happened, versus what you'd find differently if Y had actually happened, etc., then you go and *look* to see which of those side effects (if any!) you actually find when you look.
This "looking" can take different forms depending on the nature of the process being explored. In the case of some physics questions, you can build a tabletop setup of lasers in a certain configuration, or whatever, to reproduce the conditions which should act one way if hypothesis X is right, or another way if Y is correct, etc. This is a classic "experiment" in the layman's mind. In the case of medical hypotheses about what disease a patient has, you can try different drugs to see which he responds to. In some cases of geology, you can take core samples of the rocks under the Hawaiian Islands to see whether their composition and structure matches the predictions of one hypothesis, or another. In evolutionary biology, you can go check the DNA of various species to see if the pattern of differences and similarities matches the precise patterns (not just *general* patterns) which distinguish one hypothesis from another, or go find new fossils (or re-examine old ones) to see whether predicted features which no one sat down to check before match the predictions of various hypotheses, etc.
When this is done over and over and over again, and the *dozens* of different predictions of any one particular hypothesis are tested and all the predictions are found to match, and perhaps even more importantly, the things that the hypothesis predicts you *won't* find are looked for and found *not* to be present as well, then you can have more and more confidence that the hypothesis is, if not 100% correct (since you can never be *entirely* sure), very much on the right track, and is very close to being right explanation. Meanwhile, you can have great confidence that the alternative explanations which made predictions that *failed* are wrong, and can be taken out to the trash dumpster.
(Also, any one person can obviously make mistakes about reasoning out the predictions, or how to test whether the real world matches the predictions, etc. This is why science places such stress on *repeatability* -- you have to publish your data, your reasoning, your tests, etc., so that thousands of other experts can go over it with a fine-toothed comb looking for mistakes or fraud or untested presumptions, and can repeat your tests to verify that your results were valid and/or not a fluke. Also note that *this* is the kind of "repeatability" that science requires -- creationists often think that it means that the *processes* need to be repeatable, like being able to repeat the formation of the Hawaiian Islands; that's not the case. The *tests* have to be repeatable, the *verification* needs to be repeatable by anyone who cares to double-check your results, or try a new method of validating them.)
The core tenets of evolutionary biology make *VERY* specific predictions about what we should find (and what we *shouldn't* find) when we go looking at nature, and make *hundreds* of different predictions which allow multiple independent validation tests (since any one prediction might "come true" and match just by luck, even if the explanation is wrong). The tenet of common descent doesn't just predict that DNA from closely related species will "be similar", it predicts that they will be similar in *very* precise, specific ways, and that they will *differ* in other precise, specific ways. And when we examine and compare DNA, this is *exactly* what we find, *every* time we compare new DNA sequences which we had never examined before.
But wait, there's more!
Creationists often say, "but maybe DNA doesn't match all those predictions of evolutionary common descent, maybe DNA just happens to be that way because God chose to make it that way, even if those features of DNA don't seem to make sense from a 'design' standpoint (because after all, He works in mysterious ways, and He might have some Design reason for that configuration which is beyond our understanding), and it just *happens* to 'look like' the results one would expect from evolution."
Ah, but that excuse doesn't hold water.
Here's why.
The beauty of the scientific method is not only that it can decide between the hypotheses that *have* been thought of, it can also validate a hypothesis in the face of alternate *unknown* hypotheses (like the "maybe God did it instead" hypothesis).
Here's how that works.
The beauty of the "make predictions and then test them" method (especially since it's iterative -- after you do that, you make *more* predictions and test *them*, repeatedly) is that if your hypothesis (explanation) is wrong, there's *no* rational reason for an entirely unrelated explanation to "just happen" to match *all* of your explanation's 70 or 80 predictions "by coincidence", no matter *what* unrelated explanation we're talking about, including ones you haven't even thought of.
An unrelated explanation, if true, might by chance match the results of 10 or 12 of the predictions of your "wrong" explanation, but it's just ridiculous to think that (and mathematically close to "impossible" for) a process which actually works in a different manner than your speculatory explanation to "just happen" to match *ALL* the same expected results.
It's like OJ saying, "it wasn't me who killed my wife, it was some *other* guy, I don't know who... who just happened to wear my same large shoe size... and just happened to be wearing a rare, expensive type of shoe... that I just happen to have owned too... and just happened to get injured during the struggle and drop blood from his right hand... on the same night I just happened to cut that same hand shaving... and his blood DNA just happened to match mine... and he just happened to leave a glove at the murder scene that just happens to match one found in my alley... and I just happened to have received gloves just like that from Nicole as a gift and been photographed wearing them... and hairs matching mine just happened to be in the hat left at the murder scene... and fibers matching the carpet of my Bronco just happened to also... and the only hour of that day I can't account for my whereabouts just happened to be the time of the murder... and my houseguest heard thumps like someone climbing over the fence into the yard which just happened to match the time I would have had to have returned from Nicole's... But I didn't do it, it was some other guy."
Not freaking likely, is it?
When *all* of the different pieces of evidence implicate OJ, even the most unlikely and specific ones -- when they match predictions of what you'd expect to find if the "OJ is the killer" hypothesis is the correct one -- the odds of those results being "coincidental" matches with some *other* guy actually being the killer, vanish to nothingness. As attorney Vincent Bugliosi wrote in his book, in his "how I'd have prosecuted OJ" speech:
"At what point do these things stop being a coincidence, ladies and gentlemen of the jury? When you folks, as intelligent human beings using your common sense, say to yourself- 'Aw, c'mon, you've got to be kidding. It's ridiculous to suggest all of these things are just an incredible coincidence. That's not life as we know it.' That's when all this circumstantial evidence stops being a coincidence. When you people, as intelligent, sensible human beings -and that's why we selected you folks for this jury- say to the defense attorneys in this case, 'Let's stop living in a fantasy world and come back to earth.' "When a person is innocent of a crime, ladies and gentlemen of the jury, chances are there isn't going to be anything whatsoever pointing toward his guilt. Chances are there will be nothing. But now and then, because of the very nature of life and the unaccountability of certain things, maybe one thing, in rare situations maybe even two things, will peculiarly point toward his guilt even though he is innocent. And in highly unusual and virtually unheard-of situations, maybe even three things will point to his guilt, even though he is innocent. But ladies and gentleman of the jury, in this case, everything, everything, points to this man's guilt. [...] We've set forth for you a staggering number of pieces of evidence that point to this man, and this man alone, as the murderer of these two precious human beings. Under these circumstances, it is not humanly possible for him to be innocent."
This is actually an excellent description of the manner in which scientific hypotheses are "proven beyond a reasonable doubt" as well. When we test the biological hypothesis of common descent, we don't just test *one* prediction of that hypothesis. Any *one* prediction might, by chance, "just happen" to match the results of some *other* process that was "actually" responsible for the formation of modern living things. Any *one* successful prediction/test, as they say, "doesn't prove anything". The creationists are right about that. If God made living things through a process *other* than letting them evolve from a common ancestor, especially one we couldn't comprehend, it might "just happen" that this included the existence of a half-bird/half-reptile looking animal like Archaeopteryx. Just coincidence that you'd get something like that through common descent also. And maybe the non-OJ killer just happened to wear expensive, rare shoes that OJ had too. It could happen.
But science doesn't stop there.
Common descent *also* predicts that we should find shared endogenous retroviruses in the DNA related species (see the link at the top). And indeed, they are there in the DNA when we look. Match #2. Now, whatever God's beyond-our-understanding-but-not-by-common-descent method of making life, even though we *don't* know anything about it, there's *no* reason to expect that it *has* to match, down to the tiniest specific details, the several hundred exact characteristics of endogenous retroviral patterns which would have resulted from common descent. God's different-method-entirely would be very likely to have different characteristics, produce different patterns in DNA which, even if we didn't understand the reasons for those patterns, would be unlikely to match the intricate patterns produced by the common descent of past viral infections. But, gosh, they do! Does that make sense, that God's Method would *coincidentally* match the mathematically *hugely* unlikely exact patterns that common descent would have produced? It *could* happen by luck, I guess. And the real killer of Nicole Simpson might have been a one-in-a-trillion DNA match to OJ, too. It's not *impossible*.
Common descent also predicts that the number of synonymous codon SNPs between related species should be directly correlated to the time since their last common ancestor. There's no plausible "design" reason for this, by the way, since those nucleotides are functionally silent. Flip them on, flip them off, nothing changes. And when we analyze DNA, across thousands and thousands of species, the results *do* match the predictions of common descent (*and* the indications of LCA time match the fossil record -- another coincidence?) Once again, God's Method, whatever it might be, needn't "just happen" to match the results of common descent. Whatever God's design criteria, surely *some* of his choices are going to result in a design that "just happens" NOT to match the outcome of common descent in some respects or another. But here on test #3, we find that yet *again* the real-world tests beautifully match the predictions of common descent. But the creationists assure us that *this* is just a coincidence too, just like the random coincidence that Nicole's killer cut his left hand in the same place and on the same night that OJ cut his own hand too. Mere coincidence.
Item #4: Common descent predicts that synonymous protein substitutions should fit a very specific pattern across phylogenies, entirely apart from any phenotypic "design" considerations, because it has no effect on phenotype. Again, these very specific patterns *are* found every time we do DNA analysis. Again, there's no reason for God's Method to match these results, since God any other method than actual common descent would be unlikely to "just happen" to produce results that look *exactly* like the signature of common descent. And yet... that signature *is* found in DNA. Mere coincidence again? Still?
Items #5-#10,000... Well, you get the idea.
And even each "single" item above is actually *thousands* of smaller individual items in a "family" of predictions along that line of evidence, and they *all* keep matching the predictions of common descent. Coincidence?
If life is *not* the result of common descent, by what STAGGERINGLY enormous "coincidence" does God's "other method", *whatever* it might be, JUST HAPPEN to match in every large and small detail, EVERY prediction of common descent? What dumb luck for the evolutionists, eh? The evolutionists are wrong, the creationists say, but the evolutionists lucked out in the evidence lottery, because God's non-evolutionary Design criteria "just happens" to LOOK EXACTLY LIKE THE RESULTS OF EVOLUTION IN EVERY WAY.
Just by mere coincidence, of course, and it don't prove a thing. The creationists assure us of that.
Thus endeth the lesson on the scientific method, and how it is used to validate theories in hundreds of ways, large and small, and carry them far, far beyond the shaky unreliability of "simply speculations". Speculations have not passed a mind-boggling battery of tests and reality-checks. Scientific theories have -- and continue to be tested and retested every day, and adjusted as needed to keep them as accurate as possible as new evidence and test results keep being added to science's vast body of knowledge.
Epilogue
Now, for a few clarifications. In the above, I have *not* said that science could rule out the existence of God. I have not said science could rule out the *involvement* of God. (For that matter, science can't rule out the existence of Santa Claus, and neither can any other method.) Some things are impossible to demonstrate even in principle, nor does science make the attempt. What I *have* said, however, is that the scientific method can rule out certain *kinds* of explanations. In the example I gave above, what we ruled out (to a high degree of probability) was any *alternative* to common descent that did not *include* common descent. In other words, we demonstrated that *whatever* (or "Whoever") else might be involved in the formation of life, common descent *was* part of the process. God may have made us, but if He did, he did it via some variation on common descent.
The second caveat is that yeah, there's some bogus "science" out there. But for the most part, it isn't really science -- it's not stuff that has passed through the "reality-checking" filters I've described above -- it's some quacks putting on some white coats and calling a press conference to masquerade their agenda as actual science. The things that deserve the name "science" are those that have really passed the gauntlet of heavy-duty testing and restesting, and have survived all challenges thrown at it. And yes, there have been mistakes and frauds (albeit not nearly as frequently as the creationists try to claim). But science, by its very nature, is self-correcting -- eventually someone will spot the problems, and weed them out. And almost without exception, that someone is another scientist.
The final caveat to the above discussion is that in order for predictions to be testable, the world has to work in ways that are, well, *predictable*. The world (Universe, whatever you want to call it) has to be non-capricious, it has to work within guidelines, so to speak -- it can't be *deceptive*. This is often misunderstood and mislabeled as a "presumption of materialism", but it's not. As long as God doesn't play tricks on us, or act capriciously, the world is still "lawful" enough to be testable, to operate by steady "rules". As long as God/Nature/pickyourfavoritename doesn't *actively* attempt to disrupt our quest to learn how things work, the scientific method still works too. But, in exactly the same way that all the evidence of the guilt of OJ really *could* be "coincidental" if someone carefully *framed* him for the crime, it's also possible (in concept) for all of the features of life, DNA, fossils, etc. to so exactly match the results -- the predictions -- of common descent in so many different ways, but common descent *not* to actually have happened, if God *purposely* set up all the evidence in order to falsely *mimic* such results. But I don't think that *anyone* is actually willing to claim that God might be a liar, trying for some unfathomable reason to convince us of something that isn't really true.
Barring that, though (and barring newly discovered evidence overturning everything we've already discovered -- and don't hold your breath on that one), the only rational conclusion is that so many findings, in so many different independent lines of evidence, *so* closely match the *countless* predictions of common descent, because common descent *is* actually the case. Common descent has been demonstrated to be true.
If you want the bumper-sticker version of all of the above, it's the old saying, "If it looks like a duck, and quacks like a duck, and waddles like a duck, it probably *is* a duck." Likewise, when the vast amounts of evidence overwhelmingly look like common descent, then...
[A dozen creationists posting, "that doesn't prove anything" in 3... 2...]
