A considerable body of recent evidence shows that preattentive process
es can carve visual input into candidate objects. Borrowing and modify
ing terminology from Kahneman & Treisman (1984), this paper investigat
es the properties of these preattentive object files. Experiments 1-3
show that preattentive object files are loose collections of basic fea
tures. Thus, we can know preattentively that an object has the attribu
tes ''red'' and ''vertical'' and yet have no idea if any part of the o
bject is red ann vertical. Experiment 4 shows that some information ab
out the structure of an object is available preattentively, but Experi
ments 5-12 search for and fail to find any preattentive representation
of overall shape. Appreciation of the overall shape of an object appe
ars to require the binding together of local form features-a process t
hat requires attention. Copyright (C) 1996 Elsevier Science Ltd