Scientific knowledge is provisional: it is founded on empirical evidence and our knowledge of empirical evidence will always be incomplete. Thus there is always the possibility of new evidence coming along to show that some area of scientific understanding is wrong.
This leads to a common claim that science can be disregarded because its ideas are always changing. For example, the fundamentalists at Answers in Genesis say:
We agree that scientists should continually refine their views as new information becomes available, but that is precisely the problem … Evolutionary scientists have changed “common knowledge” multiple times over the past century, yet the Bible has not changed. It still clearly teaches that the universe, earth, and dinosaurs were made during a six-day period about 4,000 years before Christ.
And the complaint isn’t only from Biblical literalists. Mark Vernon, a liberal “agnostic Christian” who used to be an Anglican priest, writes in the Guardian:
There are, of course, differences between scientific and religious myths. For one thing, scientific myths are far less long-lived than religious ones. The great faiths of the world daily turn to myths that are thousands of years old and find truth leaping off the page as they read them. Scientific myths, on the other hand, do well if they last more than a century. Who today reads Newton?
Is the complaint fair? To use an analogy, science is like golf, where the “hole” we are aiming for is truth, and by a “true” scientific theory we mean one that matches the empirical reality of our universe. Nature sets the lie of the golf course and the location of the hole, and science tries to find it. So we revise our theories when better empirical evidence comes along, and each time we do that we halve our distance to the hole.
Claiming that we repeatedly halve our distance to the hole is not a claim that we can ever be sure that we have arrived at complete truth and have sunk the ball — the provisional nature of science means that we can’t — but it is a claim that as science advances our theories become closer and closer approximations to the truth.
To take Vernon’s example, Newtonian mechanics and Newtonian gravity are a very good approximation to the truth: we know that because their predictions have been thoroughly tested and they match empirical reality very well over a wide range of parameter regimes. Of course we also know that they are not perfect — they become inaccurate at speeds vastly higher than those that we are used to on Earth, and in gravitational fields vastly stronger than those on Earth. And from studying those regimes we arrive at Einstein’s relativistic mechanics, which is an even closer match to the truth, indeed so good that we have not yet found any disrepancies with empirical data.
If Vernon’s question “Who today reads Newton?” is interpreted as “Who today uses Newton’s results and formulae?” then the answer is vast numbers of us. They are taught to and used by all physicists, and for engineering, for technology, and for sending spacecraft to other planets. My own research involves looking for planets around stars outside our solar system, and requires searching large numbers of stellar light-curves to look for small dips that might signify the transit of a planet. In running Monte Carlo simulations to test whether the dip is compatible with a planet we repeatedly call computer subroutines coded with Newton’s gravity, and thus might use Newton’s formulae a million times each day.
Newton’s mechanics are good enough to be widely used today, even though Einstein’s theories match reality even better, and Vernon is wrong to disparage Newton as superseded, with the implication that any science can be disregarded if you dislike it, because one day it also will be superseded.
How do we know that, as science progresses, we are getting closer and closer to the hole? We know that because science works, and works better and better. That means that predictions made by science often come true, and engineering and technology based on science works as intended.
Is it possible that science might play a false stroke, and at times hit the ball further from the hole? Yes, in principle that is possible (scientists are fallible), but the history of science shows that these are rare and short-lived episodes, and that the next generation of scientists quickly restore progress towards the hole, guided by the empirical evidence that tells us where the hole is.
Is it the case, as some postmodernists argue, that science is a mere “social construction”, and that scientific “truth” is an arbitrary location on the golf course, decided on by social convention among scientists? No it isn’t. If you want to argue that, then try producing an alternative “socially constructed” set of predictions of future solar eclipse times, and then test them to see if they come true. Or try building a social consensus that magic carpets fly and that Boeings and Airbuses do not — and when you’ve used your magic carpet to fly to your next postmodernist conference, then we’ll take you seriously. But so long as you use the products of science (aircraft, mobile phones and laptops, in preference to magic carpets, telepathy and divination), we’ll take that as an admission that science works in a way that “other ways of knowing” do not.
How does religion, with its “truths” that “are thousands of years old” and which “leap off the page” compare? If you are a religious believer you might claim that God has written a special book, describing exactly where on the golf course the hole is and what it looks like, complete with a flag. But here the evidence is that the claims are mere social constructs; just for example, there is a strong likelihood that your religious “truths” depend primarily on the society in which you grew up.
A Bronze-age tribe will have made its best-effort, good-faith assessment of the lie of the golf course and the direction to the hole, but since they knew much less than we do they were unlikely to have been close. From that starting point one can behave as a scientist, using evidence and reason to repeatedly stroke the ball towards the hole, and eventually get very close; or one can behave as a Biblical literalist and insist that one cannot budge even one inch from that initial guess, in which case it remains just as wrong as it started out.
An insistence that the Earth is only 6000 years old (wrong by a factor of a million, which shows that our Bronze-age knowledge of our world wasn’t that great) can be maintained only by sticking ones fingers in ones ears and saying “I can’t hear you” whenever science is discussed (though of course such people will happily and hypocritically use aircraft, mobile phones and laptops).
How about the more liberal believers? They usually grudgingly admit that science is pretty good at the empirical stuff, though, as Mark Vernon does, they will over-play the provisional nature of scientific claims as a way of promoting their alternative “ways of knowing”, that they claim relate to other types of “truth” and other locations on the golf course. These other “truths”, however, are merely fond dreams, born of wishful thinking; the evidence for them cannot be seen unless you already have “faith”.
The liberal believers have never presented evidence that their religion can arrive at truths that science cannot arrive at, and their claimed other truths are just random locations on the golf course, declared as “special” by social convention and vain hopes. In the absence of evidence of these alternative truths they only have faith. In the words of the great Christian theologian St Augustine:
Faith is to believe what you do not see; the reward of this faith is to see what you believe.
Or, in the more recent words of Mark Twain’s schoolboy:
Faith is believing what you know ain’t so.
coelsblog 
