Date: Mon, 06 Apr 1998 01:56:02 -0400
From: Michael Grondin <mgrondin@tir.com>
To: crosstalk@info.harpercollins.com
Subject: Re: Patterson on stratification

Paul Miller wrote:
> catchwords in GTh could more easily argue for stratification, with the
> catchword sayings forming the core Thomas because of their linkage to
> the oral tradition making them more archaic. I don't buy this at all
> however. I am persuaded that Patterson's list is not just random.
> 
The reasoning expressed in your last two sentences seems to be based on
a false dichotomy of exactly the kind I feared might arise from taking
Patterson's "list" to be a scientifically-adequate picture of the
catchword data in GTh, which it is not.

First, let me say where I see the dichotomy, and you tell me if I'm
right: you say that you "don't buy" the stratification argument from
catchwords BECAUSE you are "persuaded that Patterson's list is not just
random". What you seem to be saying is that one can believe either (1)
that "Patterson's list is not random", or (2) that GTh is stratified,
BUT NOT BOTH. If this is indeed the dichotomy you see (and only you can
say), then I think you are mistaken. I, for one, for example, might want
to hold BOTH (1) and (2). Why not?

Of course, I may be mistaken about the dichotomy which I think you are
expressing, in which case I'm not sure what relationship is intended
between the last two sentences of the passage quoted above. But, just in
case I'm right about your thinking, let me try to say why I believe it
to be mistaken.

First, let me contrast what Patterson has actually done in GTJ with what
he might have done: he might have said something like the following:
"Here's the list of all the catchwords which occur in GTh. x% of them
occur within a single saying, y% of them occur between adjacent sayings,
z% of them occur between non-adjacent sayings, etc." (There are other
things he should have done, but let me just stop here so as not to
over-complicate the argument.)

What's wrong with just making a list of the x's and y's, as Patterson
has done? Because it illegitimately highlights only one portion of the
data, ignoring the other portion and thereby skewing our
interpretations, which, in order to be considered scientifically valid,
must be based on (a scientific sample of) ALL the data. By highlighting
just the _nearly-adjacent_ catchwords, he has drawn our attention away
from the _not-nearly-adjacent_ catchwords, thus making it appear (even
if he himself does not say so) that the dominant (indeed, the only)
organizational principle at work in GTh is the linear principle (saying
"A" connected to "B" connected to "C"). It's as if one were to spread a
bunch of marbles of various colors out on the floor, cover up all except
the red ones, then say, "Wow, look at how many red ones there are!"

Perhaps, however, you took my long list of adjacent sayings which are
NOT connected by catchwords to be an argument that the existence of the
catchwords was random. It isn't, but I'll tell you how that came about:
believe it or not, I had actually finished what I wanted to say in that
note, when it occurred to me to look more closely at Patterson's list,
which happened to be on the previous page of the passages I was quoting.
When I started making my own list, I had no idea how long it would be. I
thought it might be more than a few items, of course, but if it hadn't
been, I was prepared to merely excise that portion of my note. So I was
genuinely surprised at the extent of the list, when I came to the end.
The last paragraph of my note expresses that surprise. I guess I had
assumed that the nearly-adjacent connections were much more extensive
than they actually are. When I asked myself _why_ I should have assumed
that, the answer came back that there must be something about
Patterson's list which is inherently misleading, no matter how carefully
he explains it.

If all this is too convoluted to follow, let's bring it down to one
simple aim that I have in mind: I might want to argue, for example, that
there is an intentional connection between sayings 2 and 29, based on
the catchword 'amazed'. If you will allow me to do that, fine, I won't
quibble anymore about Patterson's methodology. But if you are going to
say that nearly-adjacent catchwords are NOT random, but that
not-nearly-adjacent catchwords ARE random, then I'm going to have to
spend some more time explaining why Patterson's presentation is not
scientifically-adequate. (Now we wouldn't want that, would we?)

Mike