Common Mistakes in Theological Research #5: The Weak Cumulative Argument

A very common mistake that researchers make is what I call the ‘weak cumulative argument’.  This involves some thesis: ‘NT author X is referring to issue E whenever he brings up words/concepts A, B, C, or D’.  This often happens in studies that are intertextually-driven.  The problem here is that often the author can find one, two, or three examples that clearly make these connections.  However, he or she ends up spending many chapters looking for more ‘subtle’ examples.  Suprise!  There are ‘subtle’ examples everywhere!  Oh how scholarship will change once people read and understand concepts A, B, C, and D with issue E in mind!

The problem here is that previously few people, if any, have observed any of these subtle connections.  What ends up happening is that the author is trying to build a cumulative argument based on two or three strong cases, and dozens of tenuous or highly speculative cases.

The reason why the author has become blind to this fallacy is that she has been so immersed in the relevant literature that she sees examples literally everywhere.  It is like when a computer screen is on the same image for days and when you turn it off, it can’t get the impression to disappear.

What can you do about it?  Well, be very careful about your methodology section.  Create an approach that is sensitive to this problem and have ways to test whether the connection you see is really there.  Secondly, be in dialogue with others and ask them (before your viva!) if they understand the connections.  Finally, be willing to label each instance.  In my own thesis, I am in danger of this problem, so each example I look at I have labelled as ‘certain’, ‘almost certain’, ‘probable’ and ‘possible’.  Obviously, the ‘possible’ ones are the least convincing, but if my overall argument is based on the ‘certain’ and ‘almost certain’, then it is nice to have other potentially correlated pieces of evidence.

There is something to say about arrangement as well.  In the book I am currently reading, his best examples (what I would label ‘certain’ and ‘almost certain’) are front-loaded (in the first two chapters), but he goes into all sorts of tenuously related more subtle examples.  This kind of deflates the argument.  You might want to sandwich by leading with your best examples, but saving some good ones for later.  You should, of course, always end strong.

This is a difficult fallacy to avoid becuase it is so subjective.  It may help to state that somewhere, but in the end your argument stands or falls on the strength of your examples.