Loading ...
Sorry, an error occurred while loading the content.

Deriving OUGHT from IS [ was Re: [evol-psych] Re: The Moral Animal ]

Expand Messages
  • Steven D'Aprano
    ... There is a subtle flaw in your argument. Suppose I am a scientist who wishes to derive the pressure that a car s tires OUGHT to be kept at: I weigh the
    Message 1 of 34 , Jan 31, 2003
    • 0 Attachment
      On Sat, 1 Feb 2003 01:49, Jeremy Bowman wrote:

      > A lot of people seem to think that Hume's distinction between "is"
      > and "ought" applies only to MORAL "ought"s. But no "ought" of ANY
      > kind can be derived from factual statements alone.
      >
      > (To "derive X from Ys" here means: "X occurs as the conclusion of a
      > deductively valid argument whose premises are exclusively Ys".)
      >
      > There is nothing terribly mysterious about it. It is simply that a
      > genuine "ought" expresses a desideratum of some kind, which may or
      > may not be realised, whereas an "is" describes an actual state of
      > affairs.
      >
      > Take a mundane example: "This car's tyres ought to be kept at a
      > pressure of 30 pounds per square inch". It's pretty obvious that even
      > if we decide to treat this as a "fact" -- so that the quoted sentence
      > is literally true -- it is a very different sort of "fact" from, say,
      > the brute fact of the tyres' ACTUAL pressure.
      >
      > No one denies that there are clear REASONS why the tyres ought to be
      > kept at that pressure. There are obvious drawbacks to driving with
      > the tyres at too low or too high a pressure. The manufacturers arrive
      > at their specifications following rigorous testing, which is enough
      > to very narrowly constrain the range of "best pressures for normal
      > conditions" -- so there is no scope for disagreement about it.

      There is a subtle flaw in your argument. Suppose I am a scientist who
      wishes to derive the pressure that a car's tires OUGHT to be kept at: I
      weigh the car, fully laden with passengers and luggage, I measure the
      air temperature and the properties of the surface on which the car will
      be driven, etc.

      But there is one factor which I haven't accounted for: the purpose of
      the car (or the REASON, as you put it). Being entirely ignorant of the
      uses of cars (perhaps I am a Martian) I do what any respectable
      scientist does: I capture a driver and analyse his brain to discover
      the answer (Martian science is terrifically advanced). Or perhaps just
      ask him "What do you want to use the car for?"

      This PURPOSE is entirely determinable, so far as the driver is able to
      articulate it. Even if the driver is unaware of his *real* motives for
      driving (perhaps he says he wishes to get from point A to point B as
      quickly as possible, as cheaply as possible, and as safely as possible,
      when his actual, hidden purposes are to blow out a tire at 120mph, spin
      out of control and be killed in a flaming wreck) then I should be able
      to use statistical analysis of the behaviour of drivers just like him
      to determine the underlying, hidden motives.

      So you can see, PURPOSE or REASON themselves become just another
      variable to be measured. Admittedly there are many potential pitfalls
      and difficulties in doing so, but as a matter of principle that is
      irrelevent.

      I don't see the advantage of the equation IS + PURPOSE -> OUGHT. As I
      see it, PURPOSE is merely another IS. It is true that if you change the
      PURPOSE the OUGHT also changes; but if you change any of the ISes
      (change the material of the tyre) the OUGHT also changes.

      Some people might argue that PURPOSE is not something that science can
      measure or describe -- after all, we can't derive it from quantum
      mechanics. But so what? Science is full of quantities that can't be
      derived from quantum mechanics, including the electron structure of the
      hydrogen atom. The coefficient of friction between the tire and the
      road is one such variable: even though it is derived from the molecular
      structures of the road and tire, their geometric shapes, and the
      electromagnetic forces between them, in practice we treat it is an
      independent variable to be measured rather than the derived quantity it
      really is.

      I will admit one thing: one cannot derive OUGHT by ignoring PURPOSE,
      even if PURPOSE is an IS. An example of faulty reasoning should show
      what I mean:

      "Since foxes eat chickens alive, we ought not worry about cruelty to
      hens in battery farms." Invalid: foxes' PURPOSE is to eat chickens,
      they don't care about their prey. Since we are not foxes, we cannot
      derive our PURPOSE from theirs.

      "Since foxes eat chickens alive, and since like the fox we care only for
      our short term benefit, we ought not worry about cruelty to hens in
      battery farms." This is a valid moral argument, although it is not one
      which many people would agree with.

      "Happy and healthy hens are more productive than unhappy hens, and since
      we wish for maximal productivity from our hens, so we ought to worry
      about cruelty to hens in battery farms." Again, it is valid moral
      argument.

      "Happy and healthy hens are more productive than unhappy hens, and since
      we wish for maximal productivity at minimal cost from our hens, so we
      ought to worry somewhat about cruelty to hens in battery farms, but not
      excessively." Again, it is valid moral argument: the profit margin is
      more important to us than the welfare of the hens, but not all
      important. (Of course this is a moral argument: morality does not
      automatically mean 'good for others'.)


      > So, given our agreed purposes, and the world being the way it is, the
      > practical "ought" for tyre pressures on a given car is rigidly fixed.
      > However, it stubbornly remains an "ought" rather than an "is", and it
      > is not derivable from any set of facts ALONE, because "our agreed
      > purposes" are an ESSENTIAL ingredient in fixing the recommended value
      > of 30 pounds per square inch. If our purposes change, then that value
      > changes too. For example, if the car is to be driven heavily laden,
      > the pressure should be higher. If the car is to be driven over soft
      > sand dunes, the pressure should be lower.

      Many other facts are not derivable either. Why should we treat PURPOSE
      as any different from them and imbue it with mystical meaning it
      doesn't have?

      The argument about IS and OUGHT is a brouhaha over nothing, but there is
      a good reason for it. People have a tendency to hide the PURPOSE in
      their moral arguments. This is especially great amongst those who
      (rightly or wrongly) suspect that their PURPOSE would not be popular if
      it were articulated.

      "Just as ants know their station in life and are content, so the working
      classes should be content with their station in life and stop aspiring
      to become upper class [since otherwise I might lose my privileged
      position in society and become a worker myself]."

      If people don't explain their own PURPOSES, we can hardly be blamed if
      we guess wrongly, can we?

      > So even a completely mundane, wholly unmysterious, rigidly
      > constrained "ought" cannot be derived from factual statements alone.

      PURPOSE is just a sub-class of factual statement, although with certain
      characteristics that make it difficult to measure objectively. This is
      not an insurmountable problem: many other variables are difficult to
      measure.

      > And so to morality, which philosophers generally do not distinguish
      > from ethics, and which is about how people OUGHT to behave, NOT with
      > how they ACTUALLY DO behave...
      >
      > Because people are complicated, and because there is a very wide
      > range of human desiderata, and because we disagree with each other,
      > the "ought"s of morality are somewhat murky. But even if there were
      > no disagreement, and the "ought"s of morality were as rigidly
      > constrained as the "ought"s of tyre pressure, they would strictly
      > remain "ought"s rather than statements of brute fact.

      Not so. If it were true that everybody in the world, living or dead,
      agreed that (say) battery hens should be looked after to the best
      extent possible, that the well-being of the hens should be put ahead of
      profits, then it would be an OBJECTIVE FACT that the welfare of hens
      comes ahead of profits and convenience and cheap eggs, and therefore
      battery farms would not exist.

      The difficulty with morality is that, because so often our PURPOSES
      remain hidden (even to us) we are reduced to arguing from incomplete
      premises. Hidden PURPOSES often include such factors as "I am
      frightened by changed, and I don't wish to be frightened", "I wish to
      tell others how to run their life", "I don't care about the suffering
      of others", "It makes me happy to see others happy, and I like to be
      happy", and so forth.

      It might be argued that PURPOSE isn't an IS because PURPOSES are inside
      our brains, while ISes are part of the objective world. But our brain
      is itself part of the objective world, so how is it possible that
      something inside our brain is not?

      It certainly is true that (unlike the Martian) we cannot, and possible
      never will be able to, derive PURPOSE from the structure of a specific
      brain. But we *can* derive PURPOSE from either subject's words ("I want
      to drive my car efficiently, without ruining the tyres") or their
      actions. (There are sources of error in both methods: people lie, and
      people act irrationally. I will consider this a flaw when you
      demonstrate any other science that has eliminated all sources of
      error.) If we accept that the subject is neither fooling themselves,
      nor lying to us, then their stated PURPOSE is as real as are cars and
      roads and pressure gauges.

      > There isn't a hope in hell of anyone ever deriving any sort of
      > "ought" from "is"s alone, and anyone tempted to think otherwise has
      > utterly failed to grasp something quite simple.

      (1) Since it IS a fact that argumentative writers always wish to be
      proven right;
      (2) and since it IS a fact that making sweeping claims without thinking
      them through often leads to error;
      (3) and since it IS a fact that error in one's argument proves that
      one's argument is wrong;
      (4) then argumentative writers OUGHT not make sweeping claims without
      thinking them through.

      (1) through (3) are ISes, and two of them are testable facts. (1) can be
      disproven if you meet an argumentative writer who doesn't mind being
      proven wrong. (2) is subject to statistical analysis by counting the
      number of sweeping claims made without thinking through that lead to
      error and comparing them to the number that did not lead to error. And
      (3) is a tautology (by definition, an argument is wrong if it contains
      an error).

      Then (4) (an OUGHT) is derived from ISes alone, and is subject to
      scientific scrutiny.



      --
      Steven D'Aprano
    • Jeremy Bowman
      ... --Trial and error (with feedback , correction, etc.) is a very effective method of arriving at facts such as: 30 ppsi optimises grip over the most
      Message 34 of 34 , Feb 6, 2003
      • 0 Attachment
        George Parrish wrote:

        > How is the tire pressure ought
        > determined? Experience with whatever
        > the tire pressure is influences the ought
        > by feedback. Echoes of ought and is
        > eventually converge at an appropriate
        > point. Understanding how that happens
        > is similar to understanding which came
        > first, the chicken or the egg.

        --Trial and error (with "feedback", correction, etc.) is a very effective
        method of arriving at facts such as: "30 ppsi optimises grip over the most
        typical range of road surfaces".

        Armed with this fact, PLUS some other assumptions that usually "go without
        saying", we would normally resolve to keep the tyres at a pressure of
        30ppsi. We might express our resolve by saying: "The tyres ought to be kept
        at a pressure of 30 ppsi".

        What are those other assumptions that normally "go without saying"? Spelled
        out explicitly, the whole "argument" might go something like this:

        1. The car runs best when the tyres are kept at a pressure of 30ppsi
        2. We want the car to run at its best
        3. We ought to do whatever helps achieve what we want
        ------------------------------------------------------
        Therefore, we ought to keep the tyres at 30 ppsi

        This "argument" just spells out in detail what is already blindingly
        obvious -- its only real virtue is that it makes things EXPLICIT:

        Premise 1 expresses an empirical fact about the car and the typical road,
        discovered by trial and error ("feedback", etc.) as above. (If the roads or
        the car were different, then 1 would be false.)

        Premise 2 expresses a psychological fact. (If we were trying to sabotage
        the car, then 2 would be false.)

        Premise 3 expresses what we might call the basic "ought" of practical
        reason. (But there are various other kinds of "ought" -- if our interests
        were not practical but moral, or aesthetic, or prudential, or
        self-destructive, or whatever, then we wouldn't accept 3.)

        Each of the premises 1-3 is ESSENTIAL. 1 and 2 are "is"s in that they
        describe facts. But 3 is an "ought", and the conclusion RELIES on it.
        Hume's complaint was that in the context of moral reasoning, too many
        writers blithely skip to an "ought" conclusion without acknowledging their
        reliance on a prior "ought" -- (the equivalent of premise 3 above). This is
        very important, because in moral reasoning the prior "ought" is a basic
        moral principle.

        No one disagrees about the basic "ought" of practical reason above, but we
        certainly do disagree about moral principles. People are prepared to kill,
        and sometimes even to die, over moral issues. It is gross negligence,
        intellectually, to fail to acknowledge the basic moral principle(s) one's
        "ought"s rely on. At best, it is deception of self and others.

        Jeremy Bowman

        *************************** ADVERTISEMENT ******************************
        Learn new skills online. Visit IOL's NEW e-learning channel and
        choose from a wide range of training courses!
        http://content.iol.ie/elearning-home.htm
      Your message has been successfully submitted and would be delivered to recipients shortly.