Nicholas Covington over at SkepticBlogs has a post that relates to the borderlands of science and philosophy, intuition and its utility. His observations are done in light of some of the blog posts by Luke Muehlhauser (who more recently went to work at the Singularity Institute and the website Less Wrong).
Philosophers have long worked with what ideas we have in our heads, the primitive notions of things such as the reliability of the senses, our reasoning abilities, that other people have minds, etc. However, there is a continuing body of growing research that shows the limitations of our intuitions and our mental faculties (some I mentioned on my appearance on the Magic Sandwich Show). Nonetheless, Nick is correct that we ought not to go all-out against intuitions having epistemic value. Though we ought to be aware of its limitations (Nick mentions the Monty Hall Problem which drives first-year statistics students and normal people nuts.) Moreover, intuitions should not be beliefs that some have called properly basic, as they can and should be subject to scrutiny. Nick concludes that we have inductive reason to take them as probably true since they seem to work rather often.
But this leaves the question of when are our intuitions untrustworthy? Should be consider evolutionary history? Richard Dawkins had commented that perhaps because we evolved at the macroscopic level we did rather than nanoscopic (extremely small) or megascopic (extremely large) we may not have the mental faculties to understand modern physics like quantum mechanics and special relativity. This implies that we would be better adapted as Newtonian thinkers, since that is the world we live in; Newton's laws and principles correctly explain most everything material in our day-to-day world, the world our ancestors would have inhabited. So this implies a potential limitation on what we can understand and what we do understand.
Unfortunately, my own research in physics education shows that college students are not Newtonian thinkers, and this has been the conclusion of PER since at least the 1980s. For example, there are the interviews done by McCloskey et al. as summarized in Scientific American back in 1983. They noted that often students seem to have something like the medieval impetus view of motion.
For example, they asked the question of you walking at constant speed in a straight line while holding a ball out and shoulder level. If you continued on an the same speed and on the same path, and you dropped the ball, would you expect to see it fall behind you, in front of you, or at your feet? Before reading the conclusion, think about what the answer is to yourself.
Don't scroll down unless you have an answer.
Have a guess now?
OK, go ahead and read on.
Well, most of the people asked said the ball would fall behind them. 49% said it would fall over the exact spot the ball was let go, and a small percentage (6%) said it would actually travel backward. Only 45% accounted for the forward velocity of the ball when it was let go. And this was asked of college students, which means students with more education than the average in the US, let alone the world.
What is going on here? Some have noted the students had a general model for how things move. One of the things most people have in their minds is that without a force things don't move. There is also the notion of a force in the direction which tends to weaken with time (like the impetus dissipating as the object traveled). More recent research argues against most people having a complete model in their minds of how things work, and Andrea diSessa's idea of p-prims tries to account for what is going on with a collection of primitive ideas about physics (physics primitive ==> p-prims). What the correct model is, I cannot say.
However, the question of where these intuitions come from is one that currently needs more research, and I am glad to be working on that currently. I shan't say too much of what I have done until publication, but what I can say I did at the AAPT meeting last month. And what I did reveal is that students claim to actually feel the sorts of forces that may account for these anomalies from reality, but those forces simply are not real. At the intuitive level, we seem to believe we are pushed around in ways that simply are not true and at variance to Newtonian (let alone modern physical ) mechanics.
This indicates that Dawkins is wrong about our adaptability to understanding the Newtonian world--we don't even get that right. Perhaps that should be no surprise since it took near 2000 years of science to come up with Newton's physics (starting with Aristotle, and then on from there). Physics teachers also have come to learn that the correct ideas are hard to get into their students, and they tend to revert to incorrect views. On the other hand, we tend not to have strong convictions about things at the extremely small level or at super-fast speeds, so perhaps the theories of special relativity and quantum mechanics are easier to learn, if not necessarily understand, since we don't experience this realm and cannot fall back onto our beliefs of what happened last time.
All this would also imply that we should be very cautious using the intuitions we build up at the level we live and project them into realms we have no experience. If our experience will betray us about moving in a circle, why think they are unquestionable with things beyond our current comprehension, such as the state of the universe before the Big Bang?
I see science as progressing by testing our intuitions and seeing that they fail, and in the process building up new intuitions via the scientific method, thus creating a coherent, tested theory. It has at least been working pretty well.
So when are intuitions trustworthy? Perhaps the best rule of thumb (which is all an intuition can be) is that intuitions are useful in situations you have had significant experience. You don't need to bring out a piece of paper and a calculator to see if a train is about to hit you based its distance and speed; you can tell you need to move, else you won't be there to analyze your intuitions!
However, this would be extremely limiting to thought experiments that philosophers use all the time. How to deal with this? That is another avenue of research that needs exploration, and I know there are people on the case. In the mean time, I have a gut feeling that this won't convince everyone to give up their gut beliefs about how the world works, from politics to morality. Oh, so much work ahead.