We present new evidence about illusory conjunctions (ICs) suggesting that their current explanation requires revision. According to Feature Integration Theory (FIT; Treisman & Gelade Cognitive Psychology, 12, 97-136, 1980), focal attention to a single stimulus is required to bind its features into an integrated percept. FIT predicts that if attention is spread over multiple stimuli, features of these different stimuli can be combined into a single percept and produce ICs. Treisman and Schmidt (Cognitive Psychology, 14, 107-141, 1982) and Cohen & Ivry (Journal of Experimental Psychology: Human Perception and Performance, 15(4), 650-663, 1989) supported this prediction. In the latter study, participants viewed brief displays containing two digits and two colored letters. Digit locations were pre-cued, and participants were instructed to prioritize the digits and to spread their attention across the region encompassed by the digits. Cohen & Ivry found that reports of one letter (the 'target') produced ICs when both letters appeared between the digits. Expanding on Cohen & Ivry's paradigm, we find that both letters do not need to appear between the digits to produce ICs. While the target letter was highly susceptible to ICs if the target appeared inside the position of a nearby digit, the position of the other letter was largely irrelevant. Our experimental results also argue that these ICs were not due to mnemonic errors occurring while the digits are being reported. Based on our findings, we propose that attention to the digits casts an attentional 'shadow' projecting towards fixation, interfering with processing of target letters in that shadow and allowing color information from elsewhere in the display to be included in the resulting percept.

Examination of certain illusory conjunction (IC) errors may provide insight into the mechanisms of object recognition when multiple stimuli are attended. An IC error occurs when a subject reports a stimulus that was not present but that combines features of target and distractor stimuli. While ICs between nearby stimuli have been frequently studied and may be related to crowding (McClelland & Henderson, 2012), the properties of ICs are less clear in cases where target and distractor stimuli are distant from each other. A series of experiments replicated and extended a representative study (Cohen & Ivry, 1989). Subjects observed a display with two white digits and a colored target and distractor letter. Stimuli were interleaved and horizontally displaced from a central fixation mark. On each trial, subjects reported the identities of the digits and then the identity and color of the target letter. The frequency of ICs was critically related to both mnemonic demands and dual task aspects of the procedure. ICs were less frequent though not eliminated when the target letter was reported first, suggesting that some ICs were related to memory errors from reporting the digits first. However, ICs were eliminated when subjects could entirely ignore the digits, even controlling for accuracy. Interestingly, when digits were attended, the relative placement of digits and letters strongly influenced errors – for example, ICs were greatly reduced when the digits were vertically offset from fixation. When the digits and letters were interleaved horizontally, accuracy was reduced when a digit was present just outside rather than inside a target letter’s position relative to fixation (the distractor letter’s position had no similar effect). This effect did not occur when subjects ignored the digits. These results suggest that attending to and encoding additional stimuli has a particular and positionally-sensitive effect on errors reporting a target stimulus. Meeting abstract presented at VSS 2013

An illusory conjunction (IC) can be defined as a perceptual error in which a subject reports a stimulus that did not appear but that combines features of the stimuli that were present. Pelli, Palomares, & Majaj (2004) noted that many IC studies use stimuli whose target-flanker proximity falls within the critical spacing for crowding. For example, Prinzmetal, Henderson, & Ivry (1995) found ICs using stimuli separated by less than 15% of the target’s eccentricity. On the other hand, Cohen & Ivry (1989) found ICs with stimuli whose spacing far exceeded typical crowding values, although they used a dual task procedure with an extra memory load. To test the importance of crowding-like proximity for ICs, we replicated Prinzmetal et al (Experiment 2, no-RSVP condition) and compared ICs when stimulus spacing was or was not within typical crowding values. We found ICs with small stimulus spacing but no evidence for ICs when the spacing was increased. Our second experiment replicated Cohen & Ivry (Experiment 3) and found ICs between distant stimuli. Following their procedure, subjects identified non-target stimuli before responding about the target. Our third experiment used the same stimuli as our second experiment, but the non-target identification task was removed. Stimulus durations and contrast were also reduced such that accuracy did not differ significantly between the second and third experiments. Without a dual task procedure, the number of illusory conjunctions was significantly less than would be expected by chance, suggesting that the ICs found by Cohen & Ivry may be related to issues of memory load and response selection. We have replicated previous IC results in finding a reduction in ICs when stimulus spacing was increased. Furthermore, our results suggest that the same spacing constraints critical for crowding might play a role in illusory conjunctions. Meeting abstract presented at VSS 2012

Computational models provide a way to quantitatively explore theories about how brains give rise to cognition and behavior. This chapter introduces some basic approaches to modeling psychological phenomena, focusing on connectionist simulations of perception and memory. First, a rationale for developing computer models of behavior and brain function is presented, along with a brief description of basic computational properties of artificial neural networks. This is followed by a series of illustrative models of object recognition, perceptual learning, episodic memory, and age-related cognitive deficits. Collectively, these modeling efforts demonstrate the utility of emulating neural mechanisms in attempts to understand both observed actions and inferred mental processes. Although much work remains before the gaps between brain and behavior are bridged, the results of computational modeling efforts so far suggest that computer simulations provide a powerful tool that ultimately may help to span this divide.

Many people have the subjective sense of being able to see more than one object at a time. However, given the large receptive fields of neurons in the later stages of the ventral visual pathway, it is unclear how two similar objects could be perceived without interfering with each other. It has been proposed that the concurrent perception of multiple objects is illusory or only explicable through mechanisms such as neural synchrony (von der Malsburg, 1999). Counter to these proposals, we develop a neural network model of object recognition capable of identifying two objects at a time given only the addition of a dorsal attentional component. This mechanism is consistent with findings from Balint's patients, multi-object tracking, and change detection tasks supporting a role for posterior parietal cortex in the perception of multiple objects. Our model consists of a ventral pathway, trained to identify objects, and a dorsal pathway, trained to transform visual inputs into potential actions. Dorsal activity emergently represented both object locations and features, consistent with studies that key parietal regions may code for certain object characteristics (e.g. Konen & Kastner, 2008). With the dorsal and ventral pathways connected during training, the network learns to utilize dorsal signals to bias ventral activity towards the correct objects while suppressing errors, allowing the correct simultaneous identification of two objects. We simulate data from illusory conjunction experiments wherein, when two objects are presented briefly, subjects often erroneously report an object which miscombines the features of the actual objects. Simulated dorsal lesions impaired the identification of two objects, with recovery of double-object identification following a similar trajectory as Balint's patient R.M. In contrast, simulated ventral lesions disrupt object identification but not dorsal functions, similar to visual form agnosia patient D.F., whose ventral pathway damage allowed her to manipulate objects she was unable to identify.

Illusory conjunctions in normal and simultanagnosic subjects are two instances where the visual features of multiple objects are incorrectly 'bound' together. A connectionist model explores how multiple objects could be perceived correctly in normal subjects given sufficient time, but could give rise to illusory conjunctions with damage or time pressure. In this model, perception of two objects benefits from lateral connections between hidden layers modelling aspects of the ventral and dorsal visual pathways. As with simultanagnosia, simulations of dorsal lesions impair multi-object recognition. In contrast, a large ventral lesion has minimal effect on dorsal functioning, akin to dissociations between simple object manipulation (retained in visual form agnosia and semantic dementia) and object discrimination (impaired in these disorders) [Hodges, J.R., Bozeat, S., Lambon Ralph, M.A., Patterson, K., and Spatt, J. (2000), 'The Role of Conceptual Knowledge: Evidence from Semantic Dementia', Brain, 123, 1913-1925; Milner, A.D., and Goodale, M.A. (2006), The Visual Brain in Action (2nd ed.), New York: Oxford]. It is hoped that the functioning of this model might suggest potential processes underlying dorsal and ventral contributions to the correct perception of multiple objects.

Complementary processing systems: A PDP model of the simultaneous perception of multiple objects Cynthia Henderson Stanford University James McClelland Stanford University Abstract: Illusory conjunctions in normal and simultanagnosic subjects are instances where the binding of visual information fails to function correctly. When presented with multiple objects simultaneously, simultanagnosic pa- tients and normal subjects under conditions of attentional loads or brief presentation times often erroneously report miscombinations of features of the objects. A connectionist model of multi-object perception examines how the concurrent perception of more than one object could occur in normal subjects and become deficient with shortened processing times. In this model, the correct identification of two objects is accomplished through lateral connections between the ventral and dorsal pathways. Lesioning of the dorsal pathway produces failures in multi-object recogni- tion characteristic of the effect of parietal damage in simultanagnosia. It is hoped that the functioning of this model might help elucidate possible processes underlying the correct solution of the binding problem in normal subjects.

Interaction of episodic and semantic memory in cued recall Cynthia M. Henderson Stanford University James McClelland Stanford University Abstract: A large body of literature suggests that memories for both episodic and semantic details contribute to our ability to recall episodic events; however, the amount of interaction between the two systems during the recall process is unclear. A cued-recall experiment tested 63 participants on the extent to which semantic associations between two mnemonic episodes affected participants recall of either episode. Participants studied 16 sentences which did or did not have elements semantically associated with another sentence, while another 8 sentences were added as fillers. Sentences semantically related to another sentence elicited significantly more responses mixing the two sentences, as compared to responses for sentences without this semantic relationship. The critical responses analyzed for this study involved cases in which participants were cued by one sentence but recalled arbitrary material from the semantically related sentence. These results suggest reciprocal interactions between memories for episodic and semantic details during recall.