In the Q&A portion of this
debate, an audience member asks Christian apologist Dinesh D’Souza, “on what
basis can you dismiss the principle of parsimony?” D’Souza replies by glibly
showing that you can't use the principle of parsimony to show that the
principle of parsimony is valid. This is very similar to the oft-used argument
(especially by William Lane Craig) that you can't use the scientific method to
show that the scientific method is valid. The problem with this type of
argument is that things like parsimony, science and the rules of rationality
are methods for validating propositions. You can't use the
scientific method to prove the scientific method, because the scientific method
is not a proposition. It's not either true or false; it's an approach, a style,
a means, of finding out which propositions are true or false.
Sunday, July 10, 2016
How can you trust the scientific method when it can't be verified by the scientific method?
Subscribe to:
Post Comments (Atom)
I posted a link to this on Reddit, and I got a comment which prompted me to write a clarification/expansion in response to it.
ReplyDeleteThis is the comment that was made:
Sorry, but this is basically on the level of "why is the scientific method true? Because it works!". How do we know the cake is good? How do we know the results are accurate in the first place? If we say "the scientific method shows that these results are accurate, therefore the scientific method is a valid way of getting accurate results", we're arguing in a circle again. It's like Christians saying the bible is true because the bible says it's true.
...And this is my reply:
In the case of a cake, it's entirely subjective, and that's fine. But as with most analogies, it's not meant to be an excruciatingly exact one-to-one correspondence. The point is just to say that it makes sense to keep using methods that get good results.
Here's a different illustration. Let's say there are two people trying to build a plane. They each have their own method for how to get their planes in a flyable condition (let's call them "the Orville Method" and "the Wilbur Method"). If Wilbur's method produces a plane that can stay in the air for a full cross-country flight, and Orville's plane can't get off the ground, then I'd say the Wilbur Method is probably the one we should keep using. Now, someone who disagrees with Wilbur's design philosophy might come along and say, "the Wilbur Method isn't trustworthy, because you can't use the Wilbur Method to prove that the Wilbur Method works!" But that doesn't matter, because the plane flies.
Here's another example. Let's say two people are standing beside each other and looking forward. One says they see a wall in front of them, and the other says they see an open space with no obstructions. If they both walk forward, and they both run into a wall, then I think we can probably draw a decent provisional conclusion that the person who saw a wall has better eyesight. We can further confirm this by snapping a picture in the direction they were facing, and if the picture turns out to be a wall, then the person who saw a wall probably has more trustworthy senses.
So, in ways like that, I contend that we absolutely can get a good impression of whether methods are effective without it being circular.