Shakespeare Electronic Conference, Vol. 1, No. 39. Thursday, 30 Aug 1990.
Date: Thu, 30 Aug 90 12:02 CST
From: This email address is being protected from spambots. You need JavaScript enabled to view it.
Ken Steele expresses some reservations over the apparent circularity in the
logic underlying research into compositor stints.  While I don't follow this
research in Shakesperean bibliography, I do think the question of circularity
is important and interesting because it arises in so many discussions of
methodology in the humanities.  (In my own specialty, Hoyt Duggan's attempts to
identify metrical innovation introduced by scribes in Middle English
alliterative poetry have met with charges of circularity.  In the field of
historical linguisitics, I think one has to be struck by the apparent
circularity of the method which governs the reconstruction of proto-languages
for which there is no surviving textual evidence.)
The problem seems to run something like this:  Suppose I can partition a text
into subsets A1 and A2 on the basis of a given feature "A" (an orthographic
habit, perhaps, or an unusual metrical pattern) which is pretty clearly
attributable to a certain cause.  I then notice that this partitioning into A1
and A2 also isolates another feature of the text, "B" (a syntactic structure,
say), the cause of which cannot be immediately identified.  How reasonable is
it to conclude that the cause of feature "B" is the same as the cause of
feature "A"?
As long as the partitions coincide exactly, I think we feel confident that such
a conclusion is justified.  But in practice, things hardly ever work out so
neatly.  (Hence Ken's interest in seeing which partitionings result in the
neatest _statistical_ probability.)  One difficulty in evaluating these sorts
of arguments seems to arise in the way counter-evidence is treated:  Should I
be satisfied if partitioning into A1 and A2 isolates 95 percent of feature "B"?
If I can explain the exeptions as being themselves rule-governed can I then use
the exception-rule in my partitioning?  How many rounds of this sort of
reasoning can I accept before plausibility is lost?
My sense is that if the process results in a relatively compact and elegant set
of "rules" for partitioning a text, my argument for partitioning will be
accepted, though doubts will remain as to whether there isn't something "at
root something . . . strangely illogical about it all."
My question is whether it is possible to distinguish between methods which are
truly _circular_ (and hence yield fatuous results since they beg the question)
and those which are merely _recursive_ (i.e., those which, although they rely
on their own results, produce a series of successively closer approximations to
"the truth")?
Eric Eliason
Gustavus Adolphus College
(This email address is being protected from spambots. You need JavaScript enabled to view it.)

Subscribe to Our Feeds


Make a Gift to SHAKSPER

Consider making a gift to support SHAKSPER.