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

Parsimony, method and all that

Expand Messages
  • gentile_dave@emc.com
    First of all, I d like to touch on one old discussion. Some time back I suggested the possibility that NT scholars might not be familiar with a Bayesian
    Message 1 of 1 , Sep 13 8:44 AM
    • 0 Attachment
      First of all, I'd like to touch on one old discussion. Some time back I
      suggested the possibility that NT scholars might not be familiar with a
      "Bayesian" method of thought, and proceed to try to figure out how
      Baysianism related to standard NT practiced like exegesis. I now believe
      I was wrong in that suggestion. At the time Bob Schacht indicated that
      he thought many NT scholars proceeded in an intuitively Bayesian
      fashion. I don't think I argued that point at this time, but now I'd
      like to fully endorse it. Having worked through the exercise once,
      http://www.davegentile.com/synoptics/Mark.html , I now believe that the
      stated method of "correct" exegesis and the Bayesian method are the
      same. That is I think all I've done is quantify standard exegesis
      procedure on my page.



      In Exegesis, I believe we take certain tools of the rational-empirical
      method out of the bag. The focus is almost exclusively on the empirical
      side of the coin. The reasons for this I'm sure are evident to all here.



      In the present discussion, I'm wondering if that summarizes Ken's
      objection to my method. I.e. he might be saying that my method does not
      look like exegesis. That's probably a fair statement. However, do we
      really want to take tools out of the rational-empirical tool kit when
      trying to solve the synoptic problem? After, all this is not Exegesis.
      Certainly the methods used in exegesis can be useful, but I would argue
      we want the full tool kit here.



      Let me describe a simple set-up where we might employ these methods. The
      example comes from Richard Bradley (Philosophy of Science, 72 (April
      2005)), I'm familiar with it because I agreed with most of the paper,
      but not his last example, as he employed it. Here I borrow it for a
      different purpose.



      Suppose we have a large grain bin, with 2 outflow pipes, A and B. We
      also have two bins where the outflow can end up X, and Y. X has 2 inflow
      pipes P and Q, and Y has one inflow pipe R. The actual structure of the
      pipe system is hidden. We can measure only the outflow and inflow. This
      is the fully empirical part. We make repeated observations, and we can
      estimate with greater and greater confidence the amount of grain going
      into each pipe. Eventually we discover that apparently A=P, and that
      B=R+Q. From this we arrive at a theory of the pipe structure. We say
      that pipe A flows directly to bin X, but pipe B splits, with part going
      to X and part going to Y.



      The split in pipe B is unobservable, so how did we arrive at that idea?
      One (somewhat artificial) way to describe the process is this - We were
      making all sorts of measurements of grain flow, treating them all as
      independent variables. We then ran in to a *rational* contradiction to
      this idea of independence. The variables did not behave independently.
      We conclude they are related by some hidden structure.



      The point of this example? Well, I am going to make a hypothesis here.
      The hypothesis would be that I don't think sticking to the fully
      empirical side of the coin, that is following the practices of standard
      exegesis, could ever uncover a hidden variable. That is if we insist on
      empiricism *only*, and completely exclude the rational side of the coin,
      we are effectively eliminating the possibility of lost sources apriori.
      I don't think we want to do that.



      Finally, regarding parsimony, I think I was too hasty in granting that
      to Ken's hypothesis over mine. He has a couple of explanatory hypotheses
      for Luke's order.



      1) Luke dislikes doublets

      2) "Close" conflation would be avoided (even if not too difficult).
      Or "close" conflation would be too difficult.

      3) And Luke is relatively uninterested (compared to 1 and 2) in an
      historically ordered account (Or Luke does not view Matthew's order as
      historical, ether because of genera, or because he does not think
      Matthew knows).



      On my side I have



      1) Luke had a saying source forged by the author of Matthew (or
      really, an authentic saying source would work just as well for this
      argument)





      We then want to ask question like -

      How likely are each of these hypotheses apriori, before we look at the
      data?

      How well are each of these hypotheses supported in the data, if at all?

      How well do the logical conclusions drawn from the hypotheses match up
      with the data they intend to explain?

      Which solution has greater parsimony?

      And of course we need to ask things like have we avoided equivocation,
      etc.



      Dave Gentile

      Riverside, IL







      [Non-text portions of this message have been removed]
    Your message has been successfully submitted and would be delivered to recipients shortly.