Despite these and similar 
remarkable findings and 
ingenious theorizing, 
questions remain about their 
interpretation. First, it is not 
quite correct to refer to the 
experiments on search as 
indicating pre-attention, for 
the simple reason that the 
subject is attending to the 
entire array of elements. The 
task requires it, although it is 
true that the subject does not 
attend sequentially to each 
separate element when pop out 
occurs. Recent research by 
Arien Mack, Christopher 
Linnett, and me and by 
Mercedes Ben Av, Dov Sagi, and 
Vochen Braun at the Weizman 
Institute of Science in Israel 
indicates that "pop out," 
grouping, and texture 
segregation do not occur if 
rigorous attempts are made to 
eliminate voluntary attention 
to the array. Indeed, 
Nothdurft’s research suggests 
that it isn’t a unique feature by 
itself or a cluster of features 
against a background of 
distractor elements that leads 
to pop out, texture segregation, 
and grouping, but rather the 
contrast or difference between 
the item or cluster at its 
interface with the background. 
Moreover several experiments 
have now shown that elements 
that would seem to be complex 
because they combine several 
features do pop out. Arien Mack 
has even found that one’s own 
first name pops out among some 
words of equal length used as 
the distractor. Apparently, 
then, many so-called features 
within a unit can be combined 
with one another without 
attention being selectively 
focused on that unit.