A Donk (1999) showed that various data patterns that have been considered a
s evidence for the existence of illusory conjunctions may be due to errors
of target-nontarget confusion, an account that challenges the mere existenc
e of illusory conjunction. In a reply, W. Prinzmetal, J. Diedrichsen, and R
. B. Ivry (2001) argued against this conclusion, claiming that some earlier
findings can be explained only when one assumes that illusory conjunctions
exist. The current article shows that Prinzmetal et al.'s claims cannot re
fute any of Donk's earlier conclusions, suggesting indeed that one can only
conclude that "illusory conjunctions are an illusion."