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

26359Re: [evol-psych] Re: Scientific justifications

Expand Messages
  • Jeremy Bowman
    Aug 3, 2003
      Having spent about 20 years of my life defending scientific realism against
      confusion of various kinds, I don't want to spend too much more time on it.
      My main claim in this discussion has always been:


      (Where '=/=' means "is not the same thing as".) Here's another way of
      expressing the same simple idea:


      A sentence is true in much the same way as a man is an uncle. When your
      sister in Antarctica gives birth, that makes you an uncle. You don't have
      to know that she is in Antarctica, or that she has given birth, in fact you
      don't even have to know that you have a sister! But as a matter of brute
      fact, if she really is your sister, and she really has given birth, then
      you really are an uncle.

      When things are arranged in a particular way, any sentence that says they
      are arranged that way is true. You don't have to believe it with any degree
      of confidence, in fact no one has to believe it at all. But as a matter of
      brute fact, if the words in the sentence really refer to those things, and
      those things really are arranged in the way the sentence says they are
      arranged, then the sentence really is true.

      So truth is a simple, "transparent" sort of relation between sentences and
      the real world. By contrast, the degree of confidence a person can have in
      one of his beliefs is not at all simple or "transparent". Whatever we call
      it -- confidence, certainty, justification, probability, whatever -- it is
      not at all like being an uncle. It is a complicated, contextual sort of
      attribute -- more like being "generous" (to pick another human
      characteristic out of the air).

      There are lots of generous people who are not uncles, and lots of uncles
      who are not generous. There are some generous uncles, but there is only a
      loose connection between being generous and being an uncle. Similarly,
      there is only a loose connection between feeling justified in believing
      something, and that belief's actually being true. You can feel very certain
      about one of your beliefs, and yet still be wrong. Conversely, you can feel
      very uncertain about it, and yet it can still be true.

      Robert Karl Stonjek writes:

      > I think most physicists believe
      > they are searching for models
      > which accurately reflect nature.

      -- I'm happy enough with that. Strictly speaking, truth and falsity are
      attributes of declarative sentences of a language. But we can stretch usage
      a bit and speak of other representational media as being more or less
      "accurate". The more accurately a model reflects reality, the "truer" it
      is. This usage meshes quite smoothly with ordinary talk of maps (say) being
      "right" and "wrong" when they more or less accurately capture features of
      the terrain. For example, early maps of North America showed California as
      an island. These maps were wrong -- stretching the word a bit, they were
      "false". They were false because as a matter of brute fact the Baja
      California is attached to the rest of America. Given that blue areas stand
      for water, brown areas stand for land, etc., in respect of this detail
      these maps expressed a falsehood rather than a truth.

      > Modelling is not the same as
      > truth.

      -- No, but being an accurate model is closely analogous to being a "true"
      model, and being an inaccurate model is very closely analogous to being a
      "false" model.

      > One can say that a model is true,
      > if we stretch the word a bit, but
      > most models are accepted as
      > accurate approximations to
      > within the parameters thus far
      > tested by experiment or
      > observation.

      -- This is an example of the sort of confusion of truth and certainty I
      have been complaining about. Any hypothesis (theory, model, etc.) is a
      guess. If it's a lucky guess, then it's true (accurate), and if it's
      unlucky it's false (inaccurate). But the extent to which a hypothesis
      (theory, model, etc.) is ACCEPTED AS true (accurate) is a measure of OUR
      CONFIDENCE in believing it. Its actual truth (accuracy) or falsity
      (inaccuracy) is a different thing altogether.

      > No-one can say that "inflation
      > theory" is true, for instance, even
      > though the basic principles of the
      > model are now generally
      > accepted.

      -- No one can say it with certainty, but surely we can say it without

      > Considering the sparseness of
      > historical knowledge that is
      > bound to plague disciplines like
      > Evolution and its various
      > branches, one can not say with
      > any certainty that any models are
      > true.

      -- Agreed, but who needs to say it with certainty?

      > Rather, they explain certain
      > phenomena, though other
      > phenomena may contradict it, and
      > they seem to predict some results.

      -- They wouldn't explain those phenomena unless we assumed the explanatory
      hypotheses were true. To the extent that models cannot be understood as
      true, they fail to explain. The basic outlines of evolutionary theory are
      not models, and they can be expressed as non-metaphorical sentences of a
      language like English, and they are literally true or false. My guess is
      that most of these sentences are true.

      > It is quite common for more than
      > one hypothesis to model the same
      > phenomena where there is a
      > sparsity of observational or
      > experimental data. Advocates of
      > each model do not generally argue
      > for the truth of their model, but
      > argue instead that their model more
      > accurately reflects the nature of the
      > phenomena so modelled.

      -- Arguing for the greater "accuracy" of one model over another is the same
      thing as arguing that it more closely approximates the truth. The word
      'true' is sometimes avoided (in physics, mostly) because these models
      consist of mathematical formalisms rather than plain English sentences. But
      many of them can be roughly "translated" into such sentences, and those
      sentences are literally true or false.

      > There are also unprovable
      > hypothesis.

      -- I would say that no hypothesis is provable, at least not in the sense
      that a mathematical theorem is provable. But again, provability and truth
      are entirely different things.

      > For instance, if we made the claim
      > that no two snowflakes are the
      > same, then this can only be true
      > until two identical snowflakes are
      > discovered.

      -- This is another an example of the sort of confusion of truth and
      certainty that I have been complaining about. The claim that "no two
      snowflakes are the same" is either true or false. We can have varying
      degrees of confidence in it. Suppose at first we are quite confident that
      it is true. But later, when we discover two exactly similar snowflakes, our
      confidence goes from "quite high" to "zero". Despite the big change in our
      confidence, its truth-vale remains completely unaffected. It is false, and
      importantly, it always was false. Before, we just didn't know that it was

      All the best -- Jeremy Bowman

      *************************** ADVERTISEMENT ******************************
      Looking for a new job? Monster.ie has thousands of jobs in Ireland to
      choose from. For info visit the IOL Career Channel in association with
      Monster.ie at http://content.iol.ie/careers/
    • Show all 6 messages in this topic