Live! Home | About Live! aaronagassi                            
Questions You've Answered
Question: (Answers no longer accepted): Are feature integration theory and Gastalt principles compatible or contradicting ?

Asked by Sag
Date: 6/18/00 6:25 a.m. PT
Answer: (1 of 2) By feature integration theory are you referring to the principles subsumed in this site:

If so, I would say that there is certainly overlap between some of the concepts espoused by Gestalt. This is based on considerable experience with Gestalt and a brief read through of the above listed site. A more careful reading could reveal contradictions in addition to the compatible concepts obivious by superficial reading of the above site. Gestalt is very concerned with the whole concept of boundaries. Good luck.

By Joel99

Date: 6/19/00 3:13 p.m. PT
 Rated: Helpful

Answer: (2 of 2) Reductionism, usually denoting reduction to an array of one or zero dimension, states that the whole is no more than the sum of it's parts, while Gestalt states that a system must be seen as a whole, distinct from the sum of it's parts. After all, there is greater complexity in the interrelation even of the same elements into entirely different forms.

Now, there are two ways to go beyond Reductionism to one or no dimensions, which is clearly so inadequate to model much of anything. One is by spatial metaphor, and another is by reduction to an array of as many dimensions as are needed to describe the complex interrelationships of the system modeled.

Gestalt interprets phenomena as organized wholes, rather than simply as aggregates of distinct parts.

"the Gestalt psychologists realised that just because you know what the elements are, doesn't mean you'll know what the configuration of elements looks like." 
"[Feature integration theory] is concerned with how the various attributes of visual stimuli are combined into single perceptions. Treisman (1988) has argued that this combination is an atttentional process."

Feature integration theory describes a perceptual process. And perceptual processes are Epistemological, and model complex systems observed. This involves pattern recognition. And pattern recognition is a conjectural process, that begins with pattern creation, and proceeds through selection by Critical Preference and then Reality Testing.

Features are coded by separate systems, and then integrated. This rather seems like Reductionism in reverse. Where as Reductionism reduces the whole down to it's parts, feature integration performs this operation in reverse. But another aspect of feature integration is selective attention flittering, in which some details are left out. Because, top down processing, simply integrating every bit of information, methodically, until the model is complete, would be too much work.

In fact, not only is much information dropped out, but certain features draw immediate attention by their presence or even by their absence.
"Hypothesis: Attention is needed to conjoin features from different maps. An "attention map" determines the location of the attentional spotlight, which can "zoom" in and out to try to conjoin features."

This is a conjectural modeling process, and sometimes, in test subjects, misconjunctions occur, when the attentional spotlight widens.

In other words, elements are sifted, selected, and then relations are drawn between them. This suggests reduction to an array of many dimensions. And that goes beyond Reductionism to one or no dimensions. Thus, arguably, Feature integration theory is a Gestalt theory.

"[Gestalt] maintains that the significance of a structured whole (e.g., in visual perception) does not depend on its specific constituent elements; thus a drawn figure still has meaning when there are gaps in the drawing."
"the Gestaltists did not provide an accurate physiological model of how the visual system processed configuruations" Feature integration theory, then, is an attempt to elaborate upon Gestalt, so that it can be tested and supported experimentally. So I really don't know why I don't find Gestalt and feature integration theory are not found mentioned in conjunction.

But perceptual feature integration is only a special case of Gestalt in the wider sense. Because Gestalt is, first of all, an Ontological statement that Reductionism is not true. That systems are structures, not merely sum of their parts. Gestalt also focuses on structural repair to solve problems, as in reintegrating parts of personality into a healthy to remedy psychological disturbance. Or the analysis of complex behavior patterns.

Gestalt can also be applied to the problem of teaching in a way that focuses on comprehensible structures, rather than in confusing pieces of information with no guidance in fitting them together, Inductively.

Feature integration theory integration also portrays perception as a Gestalt analysis of reality.

By AaronAgassi

Date: 6/19/00 4:32 p.m. PT
 Rated: Excellent

Update: 4/8/2002
In repairing the links, to make them direct, bypassing the no longer existent Looksmart cgi bin, I have come across an article more specific as to the perceptual experimental research of Gestalt, at

And it becomes the more clear precisely how Gestalt prefigures Feature Integration Theory.

Indeed, it now seems that the Gestaltists may have developed more in the way of detailed testable hypothesis than I had gotten the impression researching to write the above. And the progress in Feature Integration may constitute even more of a fine point than I had hither to construed.

Links updated again on: 2/4/2003 and then on 2/8/2007

Also see: Gestalt Learning Theory for Dummies



powered by FreeFind                                                it's private by ChangeDetection