3 ms·
Yes, I understand that. The issue is that not all disagreements are about what you call "matters of opinion" (and thus faultless [1]). (And even when they are,
by fmoralesc 10y ago
Yes, I understand that. The issue is that not all disagreements are about what you call "matters of opinion" (and thus faultless [1]). (And even when they are, one can give an approximate answer that filters out a bunch of alternative opinions on the matter: for example, it seems like a waiter in NYC should not be paid less than a certain a amount, let's say, the average it takes for someone in NYC in the person's social strata to pay rent and be able to eat and go to work each day). The issue really comes up when one takes all disagreements to take this form, which is, I think, what this thread was initially about (then the GGP asked why we should care about the facts at all).
[1]: https://www.jstor.org/stable/4545405 https://www.jstor.org/stable/4545405
- mistermann 10y agoAh, I wasn't aware there was some formal theories on this idea! Along the same lines (maybe), I've always thought an improved way to approach disagreement (breaking down "matters of opinion) is to break things down into components to better discover discrete points of contention. When you have a unsolvable dispute, you've either reached a matter of opinion, or, the disputed topic has to be broken down further. I wonder if that's the approach the book takes? I think this approach could be quite successful, assuming two reasonable people involved (a tall order nowadays perhaps).