in their responses. This was taken to imply 'grouped' or 'conjoined'.
features. Another support was that 'feature' recognition errors were much less.
frequent than 'binding' errors - seeming to support a sequential process.
An alternative view - so called Gestalt' - that was current when they wrote.
was that the whole view was absorbed and then analysed in detail as if the.
subconscious was recording everything at the same level of information and.
then processing it. This is the way, in fact that humans perceive things -.
the 'meaningful whole precedes parts and properties'[2]. This seemed to.
fail to account for the very real physical limitations of the human viewing.
apparatus. The differences in approaches have been described as either.
'scene centered' - Gestalt; or 'object centered' - Treisman.
Over the years Treisman has developed those ideas and insights both broadly.
and in detail. Most recently [Goldfarb & Triesman 2010] she has considered.
the effects of incongruent visual cues to show that effects such as having.
the word 'blue' colored in red will slow perception and may also lead to,.
firstly, complete rejection or, less likely correction of the reported to.
avoid the incongruent.
The theory has, over the years, been heavily attacked and, in the process,.
of course, it has become more complicated. There have been amendments, but.
the 'core' seems to have withstood the test of time, experiment and re-.
analysis. A very comprehensive review was published by Quinlan in 2003.
[Quinlan 2003]. He noted a number of points that have been accepted by.
Treisman:.
1) Spatial/location recognition seems to be a special case that is.
performed quickly - if not separately. That seems natural enough.
given the physical details of the eye. Tests that involve location.
are probably not, therefore, locating visual features may be a separate.
physical/mental process. This seems to have been the foundation of.
