Solid Foundations
The following is an extract from a comment I made during an ongoing discussion with fellow blogger Custard on his blog.
I give you this analogy: Before you walk into a building, you don't know how structurally safe it might be, and whether or not it might collapse on top of you. You might be 50% certain - either it will collapse, or it won't. You decide that your accumulated experience with buildings is a good guide, and enter the building. If you repeatedly return to the same building, and it doesn't collapse on top of you, your previous experience leads you to the conclusion that this building is structurally sound, and is unlikely to fall down while you are inside. You go further into the building each time you visit, increasing the risk of being trapped or hurt if it does start to collapse. You might be 80% sure, or 90%, or given long enough, 99.999999% sure. You can never be 100% sure, but you conclude that given the evidence and experience, the likelihood of this particular building collapsing with you inside is remote in the extreme.
Now, if you have carefully and clearly documented every visit, it would be the most sceptical, fearful and unreasonable of people who would refuse to enter this building on the grounds that your report does not show that the building is 100% safe, and might collapse. Of course, no-one should be compelled to enter, but this type of reaction would be unwarranted.
Taking the analogy further, if the building has been carefully planned and thoroughly tested, well designed and constructed, it will survive all manner of adversity unscathed: fires, floods, earthquakes, strong winds. By the same token, small defects might be uncovered after such events, and these deserve much scrutiny and enquiry, and ultimately some resolution.
It is probably rather obvious that the 'building' I am describing in this analogy is science, and that's exactly how I feel about it.