I am trying to use the DMM to load the sample data using the data map that
comes with the sample data. My environment is the trial VPC (virtual PC). I
think I understand the intent of the message -- if there exist two target
matches, one can't tell which the new row should match, so you declare the
new row an error. My problem is that I am getting that message when I just do
not see any existing duplicate keys. I have also verified I have no
duplication rules, so they can't be changing what "duplicate" means. Most
recently I was focusing on the Subject entity. The first two levels of data
(root subjects and children, no matter whether or not they are leaf nodes)
load fine, but the third level (all leaf nodes, if they would load)
universally get the "duplicate lookup reference" error.

Can anyone help me understand what is going on here, and help me figure out
how to load simple data?

Ah, well. The "duplicate lookup reference" errors were not spurious. Somehow,
my attempts to DMM the subject.csv file introduced at the root level
duplicates of three level-two subjects ("Mountain Bikes," "Touring Bikes,"
and "Road Bikes"). I manually deleted those and successfully processed the
file that contained only the rows that had failed to load earlier.

How those rows were inserted at the root level when each of the data rows
has a non-blank parent subject value is a mystery I am not going to pursue
right now.