Standards of Evidence

There are somethings we know about ourselves because we are human:

• We can easily be fooled (by ourselves as well as others)
• Our perceptions of the world around us are very imperfect
• Our memories of what we have seen are even more imperfect and are easily manipulated.
• People lie.

So how do we decide on how likely something is to be true?In our everyday lives we often (directly sometimes even) adopt the Missouri attitude: "Show me!".We want to see something that we can touch, feel, test, measure. If we are told that a car is worth 5,000 we like to go to the paper or online and see if others sell for that much.If the mechanic tells us we need a new transmission we might ask him/her to tell us why they think that. Do the symptoms make sense? If the work is done we might want the replaced parts. We could then, at least possibly, have someone else look at them. In fact, even a non mechanic could make some assessment of the nature of the bits.In other words, we all want some "objective" evidence. We do not want to take someone's unsupported word for it.What then is "objective" evidence:
I think it is objects, measurements and reasoning that can be carried to someone else. These things stand on their own: someone else can decide if the transmission is broken, if the measurement is correct or if the object even exists (big foot, UFO's).I would put forward that NOTHING is settled without objective evidence. The "right" religious answer never seems to get settled because it becomes a he said/she said thing for centuries.

I'm inclined to think that there is a rather too narrow view being taken of "observations". We make "observations" when we examine the scene of a crime (a fav example ). This is not examining light coming from the crime scene at the time of the crime but that is clearly not the only kind of observations we except everyday.We observe the physical objects in the scene, their condition and so on. From these we are perfectly prepared to draw very firm conclusions in some cases.With geology and paleontology and others we are able to make observations left by events in the past. We are able to draw conclusions from them and make extrapolations (predictions). These predictions may or may not be borne out by further observations. This is not, in my mind, less strong than work done in a lab. The difference is that the lab work is usually highly simplified which makes it less susceptible to confusion and other explanations. It also makes for some risk in carrying lab work to the real, messy world.With enough real world observations outside a lab we can arrive at conclusions with a high degree of certainty. We may be able to have more confidence that they apply in the real world than extrapolated, simplified lab work.

Using the example you give:
Evolution predicts changes in the makeup of the populations of living things over time. (not feeling a need to contrast this with any other view ).Does the fossil record show differing populations of life at different times in the past? How can one interpret the observation of the fossil record? I think we might need to define "interpretation" but I think that more often in the messy real world what is required is more care in considering what one is seeing. We don't control enough of the variables to get the signal clear of the noise a lot of the time. Is this "interpretation" when we sort out the mess? That is why I suggested that the real world "lab" may require a much, much larger number of data points before we can see what is going on. However, when we have them we have something that I would consider to be very competitive with lab results that have been replicated a small number of times.Using the above example; one fossil, 50 fossils, even a 1,000 observations of fossils in situ would leave a lot of room for "interpretation". However, millions paints a clearer picture with the signal very clear over the noise and is what we do to arrive at a conclusion still what we mean by "interpretation"?

- I keep waiting for someone to deliver on of these other interpretations. I think there is room to keep drilling down into more detail in the same manner. The bird reptile prediction, whale transitionals, fish-amphibian etc. are examples.ABE
I agree that in an uncontrolled non-lab environment each individual piece of data is less compelling (generally anyway - just to wish wash)Edited by NosyNed, : added a bit