Skip to content

Deriving “Ought” from “Is”

by Bill Meacham on May 3rd, 2013

Most philosophers, going back to David Hume, assert that you cannot derive normative statementsĀ (saying that you ought or ought not to do something) from descriptive statements (saying that something is or is not the case). I disagree. I have written about the topic before, and recently gave a talk to the Philosophy Club about it. I have posted a video onĀ YouTube, and you can download the presentation itself here: http://www.bmeacham.com/whatswhat/PDF/OughtFromIs1a.pdf. Here is the video:

Your comments are welcome, as always.

From → Philosophy

5 Comments
  1. Liked the video. Interesting topic. In defense of Rand’s vague ‘theory’ that morality can be derived from science. I found the lecture interesting in that it seemed to organize moral language in a concise and useful way into two camps, that of science and that of religion. BUT I don’t see a substantive way in which the information in the lecture transcends a simple claim that science is based on fact and religion is based on opinion and, specifically, cultural history. I know that seems tangential, but consider the idea that it is simply a reconstruction of Pierce’s “Ways of Knowing,” comparing his rational/scientific knowledge with a mixture of the relativistic ways of knowing. It seems simply to say that you cannot base claims on what you hear or know independent of science, or more specifically reality. This is not much of a new claim in the sense that it is what the relativists have endeavored to attack, in their minds, as easily as the objective arguments, perhaps Aristotle, prior to the culture of Science. So, in that sense, I found it organizing in a language sense, and not fruit-full in a knowledge sense.

  2. I try to get rid of the ‘oughts’ and shoulds’

    ‘If you want to get along with people, it is helpful to be honest and friendly’ is more rational, ie, you could take a survey and find out.

    Check out Albert Ellis on the matter.

  3. larry permalink

    If I tinker with your “Lassie” example of affirming the consequent, making the change from assertions about facts to assertions about wants and ought, I get the following

    1) If a being is a human, then that being is mortal
    2) Lassie wants to mortal (let’s just suppose)
    3) Therefore, Lassie ought to be human (huh? why would this follow?)

    This is quite silly. Lassie is under no obligation to be human, regardless of what Lassie wants in terms of mortality. Lassie has already achieved the supposedly desired state of mortality in a different way – by being a dog.

    This points to a systematic problem in the form of argument you’ve labeled Ethical Inference. Let’s try another example of the same form.

    1) If I hit a fly with a sledgehammer, it will die
    2) I want to kill this fly
    3) Therefore, I ought to hit it with a sledgehammer

    Well … there are easier ways to achieve the same goal. Ethical Inference looks very, very weak to me, since it ignores the fact that there can be many means to a given end.

    I can see where the accusation of affirming the consequent comes from. Affirming the consequent tries to go backwards, from the consequent of an if – then to the antecedent, ignoring the fact that there can be many reasons why a given consequent is true.

    • Bill Meacham permalink

      Just as you have to evaluate the truth of the premises in logical inference order to assess the truth of the conclusion, you have to evaluate the suitability or appropriateness of the premises of an ethical inference in order to assess the force of its recommendatory conclusion. In your first example of inference, the second premise is silly. Lassie is already mortal.

      [Your example reminds me of the Wim Wenders movie Wings of Desire, in which a disembodied and immortal angel wants to become human in order to experience things that it cannot in its angelic state. For such an angel (if its name were "Lassie") your example would indeed have recommendatory force.]

      In your second example of inference, using a sledgehammer to kill a fly, the first premise, while true, is not very workable, as you note. The strength of the recommendation is weak because of the weakness of the first premise.

      Just as a logical argument can be valid in form but lead to a false conclusion if one or more premises are false, an ethical argument can be valid in form but lead to a foolish conclusion if one or more of its premises are unsuitable or inappropriate.

      You say

      Ethical Inference looks … weak to me, since it ignores the fact that there can be many means to a given end.

      The ethical inference does not have the force of formal logic. It does not provide deductive certainty. But it does have practical import and is a useful way of reasoning, particularly if you evaluate thoughtfully the premises.

      • larry permalink

        In formal logic, a valid deductive form plus true premises yields a true conclusion. Ethical inference isn’t a valid deductive form in this sense, and the trouble springs from its form, which is closer to affirming the consequent than to modes ponens. Certainly, people do reason in this deductively invalid way all the time – and formal logic can be an aid in cataloging the possible failure modes that result.

        Formal logic can also be used as a basic framework that can then be enriched with signs that denote “X is permitted” and “X is obligitory”. All such deontic logics need “ought” in the premise to get to “ought” in the conclusion.

        http://plato.stanford.edu/entries/logic-deontic/

        The unspoken deontic premise in Ethical Inference is something like

        Agents ought to take actions that will lead to desired outcomes.

        My objections to this premise are many. Here’s just three:

        Perhaps no action is required.
        Perhaps there’s an easier way to achieve the same goal
        Perhaps what the agent (e.g. Hitler) wants is bad and wrong

Leave a Reply

Note: XHTML is allowed. Your email address will never be published.

Subscribe to this comment feed via RSS