A measurement theory of illusory conjunctions

Illusory conjunctions refer to the incorrect perceptual combination of correctly perceived features, such as color and shape. Research on the phenomenon has been hampered by the lack of a measurement theory that accounts for guessing features, as well as the incorrect combination of correctly perceived features. Recently, several investigators have suggested using multinomial models as a tool for measuring feature integration. The authors examined the adequacy of these models in 2 experiments by testing whether model parameters reflect changes in stimulus factors. In a third experiment, confidence ratings were used as a tool for testing the model. Multinomial models accurately reflected both variations in stimulus factors and observers' trial-by-trial confidence ratings.     

Holes in illusory conjunctions

To address the fundamental question “What are primitives of visual perception?” we tested the hypothesis of early topological perception employing illusory conjunctions. Through a series of experiments, which were well controlled for various feature errors, we consistently observed the illusory conjunction of holes, which are a typical kind of topological property. Subjects in our experiments perceived illusory hollow figures in which the conjoined holes underwent geometric transformations. This indicates that the holes were perceived as abstract topological entities available at an early stage, which provides further support for the hypothesis of topological perception.     

Categorization influences illusory conjunctions

Illusory conjunctions (ICs) provide evidence for a binding problem that must be resolved in vision. Objects that are perceptually grouped are more likely to have their features erroneously conjoined. We examined whether semantic grouping, determined by category membership (letter vs. number), also influences illusory conjunction rates. Participants were instructed to detect an “L” or a “7” among briefly presented character strings and to report its color. Despite high shape discrimination accuracy, participants often made color conjunction errors, reporting instead the color of a distractor character, “O”. This distractor could be ambiguously interpreted as a letter or a number. The status of the “O” was determined by other noncolored flanker characters, which were either letters or numbers. When both the target and flankers were of the same category, participants made more ICs than when the target and flankers were of different categories. This finding demonstrates that alphanumeric categorization can precede and subsequently influence binding.     

Cross-modal illusory conjunctions between vision and touch

Cross-modal illusory conjunctions (ICs) happen when, under conditions of divided attention, felt textures are reported as being seen or vice versa. Experiments provided evidence for these errors, demonstrated that ICs are more frequent if tactile and visual stimuli are in the same hemispace, and showed that ICs still occur under forced-choice conditions but do not occur when attention to the felt texture is increased. Cross-modal ICs were also found in a patient with parietal damage even with relatively long presentations of visual stimuli. The data are consistent with there being cross-modal integration of sensory information, with the modality of origin sometimes being misattributed when attention is constrained. The empirical conclusions from the experiments are supported by formal models.     

The role of attention in illusory conjunctions

In five experiments, we investigated the effects of attention on illusory conjunctions formed between features of unrelated objects. The first three experiments used a weak manipulation of attention and found that illusory conjunctions formed either among features receiving high attentional priority or among features receiving low attentional priority were not more frequent than were conjunctions formed between mixed features of different attentional priority. The last two experiments used a strong manipulation of attention and failed to reveal any evidence of true illusory conjunctions. The results are inconsistent with the feature-integration theory, which predicts that when attention is focused on a subset of items, illusory conjunctions ought to occur within and outside of the attended subset, but not between the attended and unattended items.     

Effect of feature similarity on illusory conjunctions

In four experiments, we examined whether the phenomenon of illusory conjunctions is constrained by feature similarity. Specifically, are illusory conjunctions more likely to occur between items with similar features than between items with dissimilar features? Feature similarity was manipulated in two dimensions: color and shape. Experiment 1 demonstrated that more illusory conjunctions occur between items with similar colors than between items with dissimilar colors. A similar effect was found for letter similarity in Experiment 2. Experiments 3 and 4 demonstrated that the similarity effect is still obtained for illusory conjunctions even when identification of the relevant features is near perfect. These findings introduce a new constraint in the process of feature integration. Additional implications of the similarity constraint for theories of feature integration are discussed.     

Encoding speed of visual features and the occurrence of illusory conjunctions

Earlier studies have found that people show characteristic errors when asked to observe a rapid stream of events and to name the colour of a certain letter, or the letter in a certain colour. They tend to report the letter or colour from an event later in time than the correct one. This is not true, however, if they are asked to name the colour of a digit amongst letters, without knowing in advance the identity of the digit. There is also some evidence that the errors are symmetric if the location of the target is not known in advance. Three experiments are reported which show that this symmetry results from slow encoding of the colour as compared with the digit. It is still true that uncertainty about target identity does reduce the tendency to see the target in a later than an earlier colour. This effect is, however, superimposed on asymmetries resulting from coding speed, and the effect of uncertainty of location may be due only to this.     

A model of the formation of illusory conjunctions in the time domain.

The authors present a model to account for the miscombination of features when stimuli are presented using the rapid serial visual presentation (RSVP) technique (illusory conjunctions in the time domain). It explains the distributions of responses through a mixture of trial outcomes. In some trials, attention is successfully focused on the target, whereas in others, the responses are based on partial information. Two experiments are presented that manipulated the mean processing time of the target-defining dimension and of the to-be-reported dimension, respectively. As predicted, the average origin of the responses is delayed when lengthening the target-defining dimension, whereas it is earlier when lengthening the to-be-reported dimension; in the first case the number of correct responses is dramatically reduced, whereas in the second it does not change. The results, a review of other research, and simulations carried out with a formal version of the model are all in close accordance with the predictions.     

Repetition blindness and illusory conjunctions: errors in binding visual types with visual tokens.

Repetition blindness (Kanwisher, 1986, 1987) has been defined as the failure to detect or recall repetitions of words presented in rapid serial visual presentation (RSVP). The experiments presented here suggest that repetition blindness (RB) is a more general visual phenomenon, and examine its relationship to feature integration theory (Treisman & Gelade, 1980). Experiment 1 shows RB for letters distributed through space, time, or both. Experiment 2 demonstrates RB for repeated colors in RSVP lists. In Experiments 3 and 4, RB was found for repeated letters and colors in spatial arrays. Experiment 5 provides evidence that the mental representations of discrete objects (called "visual tokens" here) that are necessary to detect visual repetitions (Kanwisher, 1987) are the same as the "object files" (Kahneman & Treisman, 1984) in which visual features are conjoined. In Experiment 6, repetition blindness for the second occurrence of a repeated letter resulted only when the first occurrence was attended to. The overall results suggest that a general dissociation between types and tokens in visual information processing can account for both repetition blindness and illusory conjunctions.     

A theory of dynamic occluded and illusory object perception

Humans see whole objects from input fragmented in space and time, yet spatiotemporal object perception is poorly understood. The authors propose the theory of spatiotemporal relatability (STR), which describes the visual information and processes that allow visible fragments revealed at different times and places, due to motion and occlusion, to be assembled into unitary perceived objects. They present a formalization of STR that specifies spatial and temporal relations for object formation. Predictions from the theory regarding conditions that lead to unit formation were tested and confirmed in experiments with dynamic and static, occluded and illusory objects. Moreover, the results support the identity hypothesis of a common process for amodal and modal contour interpolation and provide new evidence regarding the relative efficiency of static and dynamic object formation. STR postulates a mental representation, the dynamic visual icon, that briefly maintains shapes and updates positions of occluded fragments to connect them with visible regions. The theory offers a unified account of interpolation processes for static, dynamic, occluded, and illusory objects.     

Effects of luminance contrast on color spreading and illusory contour in the neon color spreading effect

The present study examined whether color spreading and illusory contours in the neon color spreading effect of Ehrenstein figures are governed by different mechanisms. In the experiment, Ehrenstein figures with colored crosses inserted in the central gaps were used. There were three luminance conditions: the luminance of the Ehrenstein figures was lower than, the same as, or higher than the luminance of the background. In each condition, 16 trials (2 sets of instructions X 8 repetitions) were conducted in a random order. Subjects were required to adjust the luminance of the colored crosses according to one of the two sets of instruction given before each trial. One was to adjust the upper and lower thresholds in the luminance of the colored crosses such that their color was seen to spread out of the crosses. The other was to adjust the thresholds such that circular illusory contours were visible. It was found that illusory contours disappeared and the color spreading remained when the crosses and the Ehrenstein figures were in or nearly in isoluminance or when the Ehrenstein figures and the background were in isoluminance. These results suggest that color spreading and illusory contours are governed by different mechanisms.     

The decomposition of visual binding over time: Neuropsychological evidence from illusory conjunctions after posterior parietal damage

Patients with Bálint's syndrome are known to make abnormal numbers of illusory conjunctions (ICs) when presented with multiple stimuli and asked to report the features of one. We used two converging procedures to assess the time course of these errors. In Experiments 1 and 2 the errors produced by a patient with Bálint's syndrome, GK, were examined as a function of when he responded. We find that ICs were present even in GK's fastest responses, but that they also increased when GK responded slowly. In Experiment 3 we varied the exposure duration of the stimuli. With short stimulus exposures GK made ICs that he was certain were correct. With longer exposures there was an increase in the number of ICs where GK expressed uncertainty. In contrast to these “uncertain” ICs, feature errors decreased as the exposure duration increased. We propose that the ICs present in GK's fastest responses, and that arise with short stimulus exposures, reflect impairments at a first stage of binding. In addition to this, “uncertain” ICs arise on trials with slow responses, and with long exposures, due to performance then being affected by impairments to a second process dependent on bound features being consolidated into a more stable representation. The role of this consolidation process is limited when responses are made rapidly and exposure durations limited. This two-stage account is discussed in relation to other accounts of feature binding.     

Rock's cognitive theory of illusory figures: a commentary.

Contrary to a previously published opinion, the 'cognitive' theory of illusory contours is rich in detail and empirical support.     

Illusory conjunctions and the cerebral hemispheres

Two visual half-field experiments tested Moscovitch’s (1979) proposition that cerebral asymmetry does not concern the earliest perceptual stages but only later processing. Subjects were briefly shown displays that included one (Experiment 1) or two (Experiment 2) types of forms differing in size and which, according to previous evidence, might lead to opposite laterality effects. Laterality effects were assessed for correct detections and for illusory conjunctions, both in terms of raw detection scores and in terms of perceptual discriminability (dr scores). In Experiment 1, displays included either rectangles or triangles. In the first case, the target was a cross; in the second case, it was a Star of David. A hemifield x size interaction was observed both on correct detections and on associated discriminability. Yet, no such interaction was obtained for illusory conjunctions or for associated d’ scores. In Experiment 2, the two types of forms were presented simultaneously, with the small ones either inside or outside the large ones. No laterality effects were observed. Some implications of these data for both hemispheric asymmetry and feature integration issues are discussed. The results suggest that early preattentive processes of feature extraction are not lateralized, whereas some integrative mechanisms, such as Treisman’s (1988) focal attention, may operate differently in the two hemispheres.     

Coaction and upward social comparison reduce the illusory conjunction effect: Support for distraction-conflict theory

This article proposes an integration of Festinger’s (1954) social comparison theory and Baron’s (1986) distraction-conflict theory of the social facilitation-inhibition effect, which successfully predicts attentional focusing in coaction when social comparison represents a distraction. Two experiments confronted participants with the illusory conjunction task (Treisman, 1998), where illusions occur because of the lack of attentional processing of central cues. If coaction, like upward comparison, is distracting and thereby enhances the attention allocated to central cues (here the target’s features) at the expense of peripheral cues (here distractors), then a reduction should be found in the illusions. Experiment 1 indeed showed a lower rate of conjunctive errors under upward comparison than under downward comparison. Experiment 2 specified that this effect was due to downward comparison effectively reducing distraction, with upward comparison only maintaining it, as compared to mere coaction.     

Spatially parallel processing of within-dimension conjunctions

Within-dimension conjunction search for red-green targets amongst red-blue, and blue-green, nontargets is extremely inefficient (Wolfe et al, 1990 Journal of Experimental Psychology: Human Perception and Performance 16 879-892). We tested whether pairs of red-green conjunction targets can nevertheless be processed spatially in parallel. Participants made speeded detection responses whenever a red-green target was present. Across trials where a second identical target was present, the distribution of detection times was compatible with the assumption that targets were processed in parallel (Miller, 1982 Cognitive Psychology 14 247-279). We show that this was not an artifact of response-competition or feature-based processing. We suggest that within-dimension conjunctions can be processed spatially in parallel. Visual search for such items may be inefficient owing to within-dimension grouping between items.     

The composition of category conjunctions

In three experiments, the authors investigated the impression formation process resulting from the perception of familiar or unfamiliar social category combinations. In Experiment 1, participants were asked to generate attributes associated with either a familiar or unfamiliar social category conjunction. Compared to familiar combinations, the authors found that when the conjunction was unfamiliar, participants formed their impression less from the individual constituent categories and relatively more from novel emergent attributes. In Experiment 2, the authors replicated this effect using alternative experimental materials. In Experiment 3, the effect generalized to additional (orthogonally combined) gender and occupation categories. The implications of these findings for understanding the processes involved in the conjunction of social categories, and the formation of new stereotypes, are discussed.