Hot tempered atheists...

(Sorry to those who don't like theoretical stuff or long posts )

Excellent! I'm glad that's cleared up. I will in the future assume that you are using the word "evidence" in your special sense. I would suggest you be explicit about the sense you are using (by using the phrase "scientific evidence" instead, for instance) so that we don't confuse people who use the word "evidence" in its normal sense. My general experience with communication has been that using words in senses different from those generally accepted (as we find them in the dictionnary, such as evidence - definition of evidence by the Free Online Dictionary, Thesaurus and Encyclopedia.) tends to lead to misunderstandings. But, hey, your choice, do whatever you like.

I'll be following this up with two points to consider. The second will clarify what providing scientific evidence amounts to, what is really being asked for, in a later post. The first follows : it concentrates on what scientific evidence is and is not. This may seem silly, but often both "theists" and "atheists" forget to clear up this important point and end up confusing other types of arguments with scientific arguments, and then don't understand on what grounds their arguments are being rejected. So, in the interest of clearing up the whole debate, here goes...

Not every bit of evidence counts as scientific evidence. A rational argument is not necessarily scientific. Scientific arguments, however, are a subset of rational arguments. Scientific arguments don't allow all the moves that one can use in a generalized rational debate. Mathematical deductions, for instance, are completely rational and yet not scientific. Arguments presented in a court of law are rational, but not scientific.

Mathematical arguments are, for the most part, deductive. I don't want to go into deductive logic because just about everybody uses it and almost nobody contests it. Point is, science uses deductive logic (and mathematics) as an auxiliary, but its essential method is inductive logic, which allows us to infer general laws (the theories) from a finite and incomplete set of observations (the data, a.k.a. "scientific evidence"). In deductive logic, the evidence (the "things helpful in forming a conclusion") are the premises of the argument. These premises (in a rational deductive argument) are not scientific evidence.

When one testifies that they saw something, this is not scientific evidence. Even though in a court of law and in our everyday lives, testimony is often considered as evidence. By law, although different countries treat it differently, it is usually considered insufficient by itself and in need of corroboration. Don't go thinking a testimony is "good" evidence, however. It is not. From a logical standpoint, it's better than nothing but only by a very very little margin (Argument from authority - Wikipedia, the free encyclopedia).

This is actually false. Testimonial mostly cannot be verified in a scientific sense. In science, an experiment is verified when it is repeated by other scientists. A testimonial is not data, it is an account. And it is usually not a claim that some theory is true. A testimonial is a report of something that happened. A testimonial in a court of law is like Tom saying, "I saw Dick shoot Harry." We cannot go back in time and verify this experimentally by repeating it somehow. We know that Dick is a human being and that human beings shoot each other occasionally, but on the grounds of mere possibility we would all be murderers (we all can, but fortunately most of us never do).

But, of course, a testimonial can be corroborated otherwise, by other evidence of a completely different nature, such as Dick's fingerprints on the gun (and this other evidence might actually be "scientific" in some sense). A testimonial may also be corroborated in a completely different way, by showing evidence that has nothing to do with the crime, but rather evidence of Tom's reliability.

Essentially, scientific evidence is all the data that is used to support a theory. Sometimes we seem to mean that the data are observations, but this doesn't mean that any observation whatsoever is acceptable. One of the great things about scientific data and scientific theories is that everyone can, at least in principle, verify them. In the case of the evidence, this means that the observations must be reproducible. And what this amounts to is to say that there are well documented procedures that anyone can follow to repeat the experiments and produce new observations, similar enough to the original observations to warrant the same conclusions. So scientific evidence isn't just any observation : it is an observation that was produced using a specific recipe. This is one of the big differences between "anecdotal evidence" and "scientific evidence".

And, incidentally, this suggests what often sets a scientific argument apart from other arguments (such as those in a court of law). In a court of law, we are trying to establish, through all sorts of different kinds of evidence (testimony, fingerprints, dna, motive, etc.), that one specific event or fact did indeed occur. A scientific theory, however, is proven (inductively) from a whole bunch of events or facts of a similar nature. As you can see, they do not have the same focus at all.

Finally, there is a very special property of scientific evidence : it only makes sense in the context of a theory. We say that scientific evidence supports a theory, and this is true. But it is also true that the theory itself defines what counts as evidence. Here's how that works, as explained with the help of an interesting historical example (citations and more info : Discovery of Neptune - Wikipedia, the free encyclopedia).

In the 19th century, "astronomers had detected a series of irregularities in [the path of Uranus] which could not be entirely explained by Newton's law of gravitation." Rather than conclude that Newton's laws were incorrect, they decided that these irregularties could be explained otherwise, "if the gravity of a farther, unknown planet were disturbing its path around the Sun." And so the irregularities became evidence of a new planet, Neptune, which was later observed through a telescope. But the irregularities could only be taken as evidence of another planet given a theory that explained how the gravitational pull of another planet could have this effect.

In other words, observations are just observations. They become real "scientific evidence" of something (in our example, another planet) only when they are subsumed under a theory and explained by it. Outside of the theory that they support they are simply "facts to be explained" (or, when the unexplained observations actually conflict with a theory, they are affectionately called "anomalies").

This last part is really really important and must be kept in mind, for the next post : when you are asked for "scientific evidence", you are also (often unwittingly) asked for a "scientific theory".

For more info on scientific evidence, see here : Scientific evidence - Wikipedia, the free encyclopedia.

There's more to say about scientific evidence and other kinds of evidence, but I think that's enough for now. I want to continue with what providing scientific evidence amounts to, what is really being asked for when someone asks for scientific evidence (and, more specifically, "scientific evidence of the existence of god").

We noted earlier that observations become scientific evidence only when they are related to a theory, and that when you are asked for "scientific evidence", you are also (often unwittingly) asked for a "scientific theory".

Let's reflect on this for a moment. Observations at once support and are explained by a theory. This sounds circular : observations support a theory, which in turn explains those observations. But this is not the case : "supports" and "explains" in science both simply mean "agrees with" the general theory. Just like when you observe a black crow, then another black crow, and so on. After coming across a large number of black crows (and no crows of any other color), you feel confident that the general statement "all crows are black" is likely correct. At this point, each one of your observations agrees with your conclusion and your conclusion agrees with each of your observations (and all subsequent observations of black crows do the same). Your conclusion is chosen precisely because it fits with all your observations. This is how inductive reasoning works, and circular reasoning is quite different : Begging the question - Wikipedia, the free encyclopedia.

But every observation of a black crow, in itself, proves nothing. It's just another black crow, it supports nothing nor is it explained by anything. It remains "anecdotal" and has no scientific status, other than as a fact to be explained. And what if you observe a white crow, what do you do? To start with, it is just an anomaly to be resolved. Do you throw away your theory (that all crows are black) or do you decree that that thing is not a crow or do you find some other way to dodge it? There are always a few options to choose from. See the Neptune example in the previous post : Uranus' anomalous orbit didn't cause scientists to drop Newton's laws. But Mercury's anomalous orbit encouraged them to do so eventually : Tests of general relativity - Wikipedia, the free encyclopedia.

So, back to our problem, providing "scientific evidence" of something means not only providing the observations themselves, but also showing with which theory the evidence agrees (how it fits with current knowledge). If there is no theory that covers this particular domain of observation, then one must provide a whole new scientific theory that does. But, unfortunately, you can't just make up a scientific theory. There is this matter of peer review and general acceptance from the scientific community to deal with. These kinds checks and balances (reproducible experiments, verification, peer review, acceptance by the community, etc.) are a great thing because they guarantee that scientific evidence and scientific theories respect a very high standard. But when you're trying to innovate, you're in for the long haul (and, in fact, given the community standard, it's completely impossible for a single person to get anywhere by themselves).

Now couple this with the fact that no currently accepted scientific theory has a place for something such as a "god". It is not an entity or a kind of entity or a construct or even a property (like a mass, or an atom, or a molecule, etc.) that any currently accepted theory allows or agrees with. Which means that you don't even have the option that your observation will be covered by some current theory. You must come up with a whole new theory of the universe that is better than the prevalent ones (or at least consistent with all of them).

Even if one does come up with a sufficiently impressive collection of anomalies, that might suggest god's existence and that anyone can reproduce reliably with experiments, no one has to agree that it is "because god". The best explanation one person can come up with isn't scientific knowledge, is it? It's not about what you can come up with, it's about what the scientific community as a whole will come up with and accept. And some people are really really ingenious at this theory making business.

Conclusion : people who asks for this kind of evidence of the existence of god either don't know what they're really asking, or they're just pulling your leg. I think it's fair to say that no one can possibly come up with what is strictly required in the short time frame of a post on these boards. And it is only slightly (infinitesimally) less implausible that they might eventually manage to convince the scientific community, given the prevalent (and very wise) feeling of "we don't go there". They stopped accepting "because god" a long time ago, and aren't about to go back anytime soon, because they can come up with another explanation if they try hard enough.

Well, the good thing for theists is, that in order to demonstrate scientifically that god does not exist, one has to go through the whole thing too. And the scientific community will not touch that claim with a ten foot pole either. So anyone who tells you they can prove scientifically that god doesn't exist doesn't really get it either. You can prove scientifically that there is no planet between the Earth and Venus. But god? Sorry, you have to define god in the context of a scientific theory before you can even start "looking for it".

Basically, gods are currently not the objects of any scientific theory.