Moving the Goalpost, or Raising the Bar, is a common informal logical fallacy in which the arguer, when presented with evidence against one of his claims, redefines his claim without acknowledging the validity of the evidence and counterargument. In other words, the arguer doesn't like what he hears so he simply changes what would satisfy the argument. In doing so, it can make any claim at all vacuously true and invulnerable to reasoned disproof.
This is Part 9 in a series about Logical Fallacies. We are going through one fallacy at a time. There are many types of fallacious arguments. I’m going to try to explain them with examples then find ways to help you refute those arguments when they occur. Please comment or email if there’s a particular fallacy you want me to tackle, or if you have success with refuting an argument using a good technique you can share.
Antagonist: "Evolution is clearly impossible; no life form can change"
Protagonist: "Um, livestock breeders do it all the time. Where do you think hybrid roses come from?"
Antagonist: "Well, that's just microevolution. You breed a new rose, it's still a rose. What you can't do is breed a new species."
Protagonist: "Actually, we can and have. There's lots of examples of observed speciation.
Antagonist: "Yes, but you still just get another variation of the same kind; you never get a completely new type of animal. You can't breed a dog and get a chicken."
The key to understand this fallacy is to understand what a claim under discussion actually means. In most cases, the actual "claim" is a relatively broad and perhaps ill-defined one. In most cases, the person making such a claim will have an intuitive, informal idea of what he really means, but cannot necessarily articulate the exact evidence upon which he bases his idea. Some concepts are hard to articulate and even harder to demonstrate, but it may nevertheless be real.
On the other hand, "moving the goalposts" can also be a sign that the claimant has made up his mind and is impervious to evidence. If he is convinced, for example, that a pattern exists, any single counterexample can be dismissed as unrepresentative.
Exceptions to the Rule:
"Moving the goalposts" can be legitimate when used to make more explicit exactly what is meant by a given claim. When the proposed amendment to the claim is more accurate and useful than the original claim, then moving the goalposts is simply an intelligent response to valid criticism.
Also, claims of "moving the goalposts" often degenerate into mere semantic quibbles when the overall meaning of the statement is clear. For example, a claim that "all scientists today accept the theory of gravity" is probably false in detail. If you spent the next five years searching assiduously, you could probably find a single person, somewhere in the world, with training in science who holds a contrary opinion in the teeth of near-universal disdain and all standards of evidence. Focusing on the lone holdout adds legitimacy to an otherwise nonexistent controversy.
I think we can say that it's probably best to avoid certain words like always, never, all, none, etc. These can easily be nitpicked as incorrect. In the first example, I think I stated that it's always better to tell the truth (I can't remember how I worded it). I really should have said, in most cases, it's best to tell the truth.
How to Refute it:
Call them on it. Tell the person they are changing the parameters of the question. Ask for one clearly defined question to discuss. Make sure the terms are clear up front.