This creates the mistaken implication that there is some need for an affirmative belief to sum up to 100%, and I think it improperly relabels "uncertainty" as "faith."
My agreement with this statement goes beyond simply voting you up. While I can see people having "faith" in placeholder evidence for something not yet found, I do not think this is a useful description when talking with people of faith.
In terms of easing the discussion between people of "faith" and people of "evidence" I do encourage finding a way to translate "evidence" into the language of faith. I just think the OP was a little off. My first attempt would be something such as, "My faith comes from induction and past experience." When they say, "Me too!" you now have a foot in the door to talk about evidence without using the word.
If they say, "My faith comes from fuzzy feelings and what my elders taught me," they are being rather honest about their faith and could probably have an intelligent conversation with you about the subject. Namely, you can contrast and compare how accurate fuzzy feelings and elders were in your life and see how they respond. Again, you can talk about evidence without using the word.
If they say, "Faith is believing in the face of uncertainty" the conversation can drift into "completing the job of evidence," which is what I think the OP was talking about. If evidence gets you 90% of the way there, but you are acting as if it were true, than there is some amount of "faith" involved. But, in my opinion, there should be a 90% faith in the evidence you found, not 10% faith in the evidence you didn't. They may have 10% faith in the non-evidence, but I would argue that this is where the rationalist line should be drawn. This is more obvious when there is 99% that is missing.
Often when talking with people of faith about the issue of atheism, a common response I get is:
Any atheist who has ever talked to a theist has likely encountered this line of reasoning. There is a distinction to be made between the glorifying of faith as evidence of truth, as theists do, and the desire to become less faithful in our understanding, as the rational truth seeking person does.
As has been discussed previously, realistically P = 1 is not attainable, at least for the foreseeable future. Nor is P=1 necessary to be comfortable in the universe. Richard Feynman and Carl Sagan among others, feel the same way. We know that heuristics “fill the gaps” in knowledge of recognizable scenarios such that we can comfortably demand less evidence and still come to a reasonable conclusion about our surroundings in most cases. We also know the myriad of ways in which those heuristics fail. Much has been written about uncertainty in the decision making process and I would imagine it will be the focus of much of the future research into logic. So, I have no problems concluding that there is some level of faith in all decision making. How much faith in a claim am I comfortable with? The way in which I have been thinking about the different levels of "faith" recently is as a sliding bar:
On the left side, natural evidence is indicated in blue which is comprised of data points which indicate to the observer the natural proof of the argument. On the right, faith is indicated in red which is comprised of either counter evidence for the claim or a lack of information. The bar’s units can be anything (probability, individual records) so long as both sides are measured in equal parts.
A few examples:
Claim: Natural and sexual selection as described by Charles Darwin accurately describes the processes which develop biological traits in separate species.
I agree with this claim with 99% natural evidence in the form of observed speciation and fossil and geologic discoveries and accept 1% faith because, well, anything can happen:
Claim: The price of Human full DNA sequencing will fall below $1000.00 before 2012.
I agree with this at 80% natural evidence based on the projection that the price will continue to fall at the same rate that it has and accept 20% faith based on the fact that predictions failed for 2008 and 2009
You get the idea.
I started noticing a few years back that most rational people, and in general people within scientific and analytic communities, generally needed to slide their evidence meter pretty far to the right before they would accept a premise, even tentatively. After all, even frequentists, as flawed as their methodology may be, still chose 95% as the conventional level of significance. Conversely with a meter far to the left, demanding much faith, most rational people I have found choose to reject those claims.
It’s not a perfect method, nor is it terribly rigorous. However I’m not interested in more rigorous methods because I am interested in bridging the gap between the rationalist and, let’s say, those who have a propensity of demanding lower levels of evidence for beliefs. I think the scale method is a grounding point for discussion between two people who hold different views on a subject. I am of the mind that this helps the average person who has never heard of probability theory grasp the idea and why one person's cause for faith is another person's cause for natural evidence. This way, the faithful can discuss why the incorruptibles are a miracle and proof of faith and another can point to the evidence of embalming, mummification or preservation like burying environment in which they were buried.
One question I have myself about the methodology: would the most rational individual restrict themselves to two scales for all claims; one for acceptance and another for rejection?
Any refinements are welcome – it’s an open source methodology.
My scale of trust in my scale method is 50% evidence and 50% faith.