X-Message-Number: 10195 Date: Fri, 7 Aug 1998 07:56:46 -0700 From: Tim Freeman <> Subject: Really can't derive "ought" from "is" In cryomsg 10177, says: >3. Philosophers often claim it is impossible to derive an "ought" from an >"is." (I think Brian agrees with that.) I can't quickly prove my claim that we >can always derive "ought" from "is," but I can quickly disprove the >philosophers' claim that we never can--because I only need one counterexample. You've misclassified "is" and "ought" in the example that follows, so you really didn't jump the gap between "ought" and "is". Generally speaking, to solve a problem it's best to trim hair from it rather than adorn it with additional hair. In this case it would work better to give the commonly accepted inference rule that starts with "is" and ends with "ought", rather than give a hairy example, but I'll address the example: "Is" statements are statements about the world -- statements about configurations of the universe that may be true for some configurations and false for others you could imagine. The statement claims that the actual universe is one of the configurations for which it is true. "Ought" statements are moral imperatives. I can't give more details about that because I don't know them; "ought" statements seem fairly meaningless to me, but not to other people, so I regard them as social phenomena rather than as statements. But surely you know what I'm talking about, so let's continue. Applying this distinction to your example: >Consider an ordinary person in ordinary circumstances. He wants to maintain >good health for an extended period. Evolution is a noisy channel connecting survival to behavior. The noise is nontrivial, so even though most ordinary people will want to maintain good health for an extended period, some others will give priority to incompatible things. So if you are talking about *all* ordinary persons, you're making an "is" statement that is simply wrong. To continue, let's suppose you have one of the ordinary persons in mind for which the statement is true. This is an "is" statement. >To do so he needs to eat a reasonably well balanced diet. Another "is" statement. So far so good. >Hence he "ought" to do so, This is ambiguous. Maybe you mean: As part of pursuing his goal of maintaining health, he ought to eat a reasonably well balanced diet. Despite the appearance of the word "ought", this is an "is" statement, making a statement about the world: people generally have to eat a well balanced diet to maintain health. This statement is true, but it isn't an "ought" statement, so you failed to jump the gap. Or maybe you mean: It is morally imperative for him to maintain a reasonably well balanced diet. This is an "ought" statement, but it's a non-sequitur that isn't supported by the prior statements, since we don't have any commonly agreed upon inference rules that derive moral imperative from anything else. There will be people who would prefer to remain healthy, but want to eat mostly junk food, and neither you nor I can prove that they "ought" to be doing something different, like translating their health goals into reasonable action or even reasonable beliefs or reasonable goals about diet. >and [he ought to] want to do so [that is, eat a reasonably well >balanced diet]. Several plausible meanings here. First: It is morally imperative for him to want to eat a reasonable diet. This is an "ought", but isn't supported by the prior discussion. For another possible meaning, we could read "ought" as in 'If you push the "on" button, the computer ought to boot', which really means 'The speaker believes that if you push the "on" button the computer will boot': Ettinger would expect him to want to eat a reasonable diet. This is an "is" statement about Ettinger, and is either true or false depending on the beliefs of Ettinger, so it doesn't jump the gap. It's interesting to look at the assumptions that may lead to the expectation. One possible assumption is that people are rational; this assumption is false since people use a bunch of randomly firing neurons to do their thinking with, among other reasons. Another possible assumption is that evolution universally causes behaviors that promote survival or at least reasonable pursuit of stated goals; as I said earlier, evolution is a noisy channel, so that assumption is false too. >This higher level >or derivative value or goal--to eat a reasonably well balanced diet--is >objectively validated because it tends to further a more fundamental value, to >maintain good health. This is an "is" statement about cause and effect -- eating a reasonably well balanced diet helps to maintain good health. It's a shame that words like "should" and "ought" have so many different meanings if you look them up in the dictionary. People get the meanings confused with each other and it often makes them do strange and useless things. (I am making a general statement here, not a vague statement about anything I've seen Ettinger do.) >...if anyone questions this position, his challenge is to >offer something different. To my knowledge, no one has ever suggested an >alternative that makes the slightest sense. The alternative is that the world really doesn't tell you what to do, so you decide arbitrarily among the choices that seem to fit whatever biases you start the game with. It's a perfectly logically self-consistent thing to do, and I don't see any non-arbitrary alternative. My arbitrary choice for the time being is to try to survive until the issue of survival becomes moot one way or the other, and at that time to take a different arbitrary choice. For evolutionary reasons one would expect this to be consistent with the biases I'm starting the game with, if the noise in the channel doesn't get me. (Given my experiences with attempting calorie restriction, the noise in the channel may be getting me. But that's another issue.) -- Tim Freeman http://www.infoscreen.com/resume.html Web-centered Java, Perl, and C++ programming in Silicon Valley or offsite Rate This Message: http://www.cryonet.org/cgi-bin/rate.cgi?msg=10195