You're all familiar with the disagreement hierarchy, right? Actually, I'm not sure how helpful it is most of the time, as I feel like a lot of the arguments I see (at least on the internet) consist of people arguing at cross-purposes rather than actually disagreeing with each other. Nonetheless, I would like to suggest two revisions to it.
Revision 1: Add level DH4.5: Nonconstructive refutation.
The archetypical example of refuting an argument is finding a hole in it -- "Your inference of P is unjustified given only what you've established so far." (Or, better yet, "Your inference of P is unjustified given only what you've established so far; indeed, here is an example where what you've established so far holds, but P does not.") But it's possible to show an argument wrong without actually finding a hole in it. The classic example is showing that an argument proves too much. If an argument proves too much, you can conclude that it's wrong -- but you still don't necessarily know exactly why it's wrong. It's still a form of refutation and should be above counterargument, but it's not as good as a constructive refutation.
Revision 2: Replace DH6, "Refuting the central point", with "Refutation and counterargument".
"Refuting the central point" doesn't really strike me as qualitatively different from "refutation". Honestly to my mind, if you're refuting some peripheral thing, that hardly even counts. When I argue I like to spot the other person lots of points because I want to get to the central disagreement as quickly as possible; arguing over peripheral stuff is mostly a waste of time. Of course, sometimes peripheral stuff becomes central later, but you can always un-spot a point.
Anyway, point is, what is qualitatively different is refuting and counterarguing. If you only refute but you don't counterargue, all you've established is that the other person's argument is wrong -- not that your own position is right! Refutation does not automatically include counterargument, and I think this is worth singling out a separate higher level.
(Sometime, I really need to get around to writing "Harry's guide for how to have an argument well"...)
-Harry
Revision 1: Add level DH4.5: Nonconstructive refutation.
The archetypical example of refuting an argument is finding a hole in it -- "Your inference of P is unjustified given only what you've established so far." (Or, better yet, "Your inference of P is unjustified given only what you've established so far; indeed, here is an example where what you've established so far holds, but P does not.") But it's possible to show an argument wrong without actually finding a hole in it. The classic example is showing that an argument proves too much. If an argument proves too much, you can conclude that it's wrong -- but you still don't necessarily know exactly why it's wrong. It's still a form of refutation and should be above counterargument, but it's not as good as a constructive refutation.
Revision 2: Replace DH6, "Refuting the central point", with "Refutation and counterargument".
"Refuting the central point" doesn't really strike me as qualitatively different from "refutation". Honestly to my mind, if you're refuting some peripheral thing, that hardly even counts. When I argue I like to spot the other person lots of points because I want to get to the central disagreement as quickly as possible; arguing over peripheral stuff is mostly a waste of time. Of course, sometimes peripheral stuff becomes central later, but you can always un-spot a point.
Anyway, point is, what is qualitatively different is refuting and counterarguing. If you only refute but you don't counterargue, all you've established is that the other person's argument is wrong -- not that your own position is right! Refutation does not automatically include counterargument, and I think this is worth singling out a separate higher level.
(Sometime, I really need to get around to writing "Harry's guide for how to have an argument well"...)
-Harry
